Evolution of AGN

Although the simple dust obscuring torus accounts for the data remarkably well, it is not generally believed that AGN are static, non-evolving objects (see review by Dopita 1997). During the period of time that a galaxy is undergoing Seyfert activity the presence of an obscuring torus plays a crucial role in determining whether we classify it as a Seyfert 1 or Seyfert 2. But the galaxies must evolve in the sense that there is a stage of time prior to the onset of nuclear activity and a stage after which nuclear activity has subsided. It is known that interactions between galaxies provide the means for funnelling gas towards the central regions of galaxies (eg. Barnes & Hernquist 1996; Bekki & Noguchi 1994), and this may feed an existing massive black hole and/or create the onset of nuclear starbust activity which could eventually lead to the formation of an AGN (eg. Planesas, Colina & Perez-Olea 1997; Colina et al 1997; Barnes & Hernquist 1991; Norman & Scoville 1988). Stellar bars, whose origin may be either internal or external, may also generate an inflow of gas towards the nucleus thereby leading to the onset of nuclear activity (eg. Mulchaey & Regan 1997; McLeod & Rieke 1995).

Compact Radio Cores in AGN

Recently, Heisler et al. (1998) found that the IRAS-selected sample of 60 m Peakers have radio core luminosities which are intermediate between ``normal" Seyferts (i.e. FIR-selected Seyferts chosen without regard to FIR spectral energy distribution) and radio-loud ellipticals (Figure 5). The Seyfert 60 m Peakers (60PKs) follow the same relationship of ``core to total" flux as radio elliptical galaxies (Slee et al. 1994) (Figure 6). Thus the correlation between core and total power for radio ellipticals extends to Seyfert 60PKs with radio luminosities as low as 10 W Hz, and whose optical morphologies tend to be peculiar (i.e. tidal tails, rings, shells, double-nuclei) (Heisler & Vader 1994), but whose near-infrared morphologies are well fit by the classical r law for elliptical galaxies (Heisler, De Robertis & Nadeau 1996). Thus, the compact radio cores in the centres of 60PKs ``know" about the total radio emission in the galaxy. Compact radio cores are presumably associated with AGN, while the total radio emission is traditionally associated with star formation throughout the galaxy. However, it is noted that the radio emission in 60PKs is not like that of classical radio galaxies. It is unresolved on scales of kpc (Vader et al 1993), and high resolution radio mapping is crucial for investigating the source of the total radio flux (i.e. radio jets or star formation).

 
Figure 5: a) The distribution function of radio cores for Seyfert 60PKs. The median is log(P=22.1)
b) The distribution function of radio cores for IRAS Seyfert galaxies without restriction to FIR spectral energy distribution from Roy et al 1994.. The median is log(P=20.9)
c) The distribution function of radio cores for radio elliptical galaxies from Slee et al (1994). The median is log(P=23.1)

 
Figure 6: Relationship between core radio power and total radio power. The best fit line is given by P P

Far-Infrared - Radio Correlation

There exists a tight correlation over a scale of several magnitudes between the radio and FIR fluxes for starburst and spiral galaxies (eg. de Jong et al 1985). Generally radio-quiet Seyfert galaxies show a larger scatter to this correlation (Norris et al 1988), while Seyferts with strong radio cores do not fit the correlation (Sanders & Mirabel 1985; Wilson 1988; Baum et al 1993; Heisler et al 1996; Roy & Norris 1997). The evolutionary significance of these radio-excess galaxies (relative to the Far-Infrared - Radio correlation) has recently become an exciting controversial issue. It may be that these radio-excess objects are young radio galaxies in which the jet has yet to blast through the interstellar medium of the galaxy. Such a scenario has been proposed for the compact steep spectrum (CSS) and Giga-hertz Peaked Spectrum (GPS) radio sources (eg. Bicknell, Dopita & O'Dea 1997a). Recently, McGregor, Kewley & van Breugel (1997) have identified a large sample of such intermediate radio-loud active galaxies, and find they have optical spectra with the signature of a post-starburst AGN. McGregor & van Breugel are in the process of testing observationally whether there is a link between radio-excess galaxies and CSS sources, and thus whether they represent a transient phase in the evolution of AGN.

Evolution of radio jets in AGN

Bicknell et al (1997b) have recently examined the question of whether the narrow line regions in Seyferts can be powered by radio jets. This has important implications for the radio-excess Seyferts. The well known correlation between radio power and the luminosity of [OIII]5007 in Seyfert galaxies (eg. Whittle 1985) points to a connection between radio-emitting and emission-line plasmas. It is well known that Seyfert galaxies containing luminous linear radio sources tend to have broader forbidden emission line widths indicating an additional acceleration mechanism which presumably involves the interaction of a jet with the circumnuclear interstellar medium (eg. Whittle 1992). When this is combined with the observation that the radio axes in Seyferts correspond to the axes of the ionisation cones, it strongly suggests that the radio and optical emission are causally connected (eg. Wilson & Tsvetanov 1994). Based on this Bicknell et al (1997b) assert that the mechanical energy and momentum transported by the radio-emitting plasma is responsible for the excitation of some fraction of the narrow-line region. They calculate that in order to the NLR emission in Seyferts to be powered, at least in part, by radio jets, requires jet energy fluxes of order 10 ergs s, and a smaller ratio of radio power to jet energy flux than is typically assumed for radio galaxies. This lower value is attributed to the younger ages of Seyferts compared to radio galaxies. Many of the 60PK Seyferts lie in a region of the radio-[OIII] diagram which is intermediate between radio-quiet Seyferts and radio-loud ellipticals (Bicknell et al 1997), which may be interpreted in terms of a transitional stage of evolution from radio-quiet to radio-loud due to the rapid growth of the central black hole as a result of a major merger. This interpretation is consistent with observational results (Heisler & Vader 1994, 1995; Heisler, De Robertis & Nadeau 1996, Heisler et al 1998) that indicate Seyfert 60PKs are intermediate between radio-quiet Seyferts and radio ellipticals and represent galaxies which have recently undergone a major merger event. High resolution radio mapping of objects such as 60PKs will clarify the importance of jets in Seyfert galaxies. IC 5063 is a Seyfert 60PK which is reasonably close and has been recently mapped in the radio, and I discuss it below.

The Case of IC 5063

IC 5063 is classified as either a peculiar elliptical or SO, and contains a series of dust lanes which strongly suggest a recent accretion event has occurred. The galaxy is a 60 m Peaker, and contains a hidden broad-line region discovered via spectropolarimetry (Inglis et al. 1993). The galaxy also exhibits a cross-shaped ionization region, which Colina et al. (1991) suggest is the edges of an ionization cone illuminated by the central AGN. Figure 7 displays our discovery of a linear triple of radio emission at 3 cm and a possible large-scale bow-shock at 6 cm in IC 5063 (Bransford et al 1998). The components of the triple are separated by approximately 3 arcsec corresponding to a linear scale of 500 pc. We interpret the linear triple source as the ejection of radio plasma from the nucleus of the galaxy. The central knot of the triple radio source is centered on the peak in the K-band continuum, suggesting that it is associated with the nucleus.

 
Figure 7: Large image shows 6 cm radio map of IC 5063 where feature towards the east is believed to be a large-scale bow-shock, and inset image shows central region of IC 5063 at 3 cm where linear triple radio source is clearly detected.

Triple radio sources, similar to that seen in IC5063, are commonly associated with Seyfert galaxies and are usually interpreted as being evidence for the ejection of plasma from the AGN. We have detected near-infrared line emission from the three knots (equivalent widths, for example in 1-0 S(1) molecular hydrogen, are 29 and 25 for the outer knots and 6 for the central knot). This is consistent with the central radio knot being coincident with the brighter K-band continuum nucleus, and the strength of the emission from the outer radio knots supports the view that their associated plasma is interacting with the ISM surrounding the nucleus. The H emission is stronger in the outer knots than the nucleus, which is consistent with the idea that the outer knots have been ejected from the central source. Recently a strong blue-shifted (-600 km/s relative to the galaxy) HI absorption line has been seen against the radio sources by Morganti, Oosterloo & Tsvetanov (1998). It was suggested by these authors that the absorption is evidence of outflow from the nucleus. We speculate that the HI absorption line is associated with entrained HI connected with the ejection event.

Because of the very dusty nature of this galaxy, we believe that a clear understanding of the dynamics of the inner regions can only come from IR spectroscopy. Planned observations with MPE 3D imaging spectrometer at the AAT will be used to provide a map of near-infrared emission lines across the radio sources and allow us (a) to resolve possible shock structures associated with the radio components, (b) to map the velocity field of the gas, and (c) to test the idea that the 21cm HI absorption results from gas entrained in the ejected radio plasma. With the velocity resolution provided by 3D it should be possible to determine if the emission associated with the radio triple shows any evidence of outflow, and in particular whether the radio knots are associated with the HI absorption outflow.

© Copyright Astronomical Society of Australia 1997