6.2 Implications for the state of the torus
6.2.3 Other aspects of AGN unification
Dust tori at low luminosities In section 2.2.3, I noted that, incidentally, the two very different torus models by Elitzur & Shlosman (2006) and H¨onig & Beckert (2007), both predict that AGN tori are not stable at bolometric AGN luminosities below Llow ≈ 5·1042
erg s−1. From this prediction we would expect low luminosity AGN to appear differently in the IR. The naive expectation is for the LMIR/LXratio to decrease as the IR emitting
torus is missing. In fig. 5.32, we indeed see a weak tendency for objects with LX. 5·1042
rather than a decrease. I should note that none of these low luminosity objects belongs to the well resolved objects discussed in section 5.3.1 which means that their deviation from the correlation could simply be caused by contamination of the observed MIR flux. Moreover, as we can see in fig. 6.3, the deviation of these low luminosity objects does not exceed the scatter among other contaminated objects.
An interesting test case for the behavior of low luminosity AGN is the well studied LINER NGC 4579 (highlighted in fig. 5.32). Contini (2004) analysed emission line ratios in the optical spectrum of this object and came to the conclusion that the MIR continuum of NGC 4579 is dominated by thermal emission of the shock heated ISM in the vicinity of the AGN. First of all, this supports the hypothesis that the deviation from the correlation is caused by contamination – in this case emission from the ISM. Nevertheless, if we would find this kind of contamination in all low luminosity AGN, this would imply a physical connection between the absence of a obscuring torus and gas outflows from the AGN. At the moment, however, our data is not sufficient to constrain the properties of this class of objects. Future observations are required to settle this question.
The role of LINERs LINERs are one of the more enigmatic classes of AGN. For a long time, it has been unclear whether they are predominantly powered by AGN or starburst activity. Some LINERs – among them NGC 4579 – exhibit broad optical emission lines, indicative of accretion activity in their centres. Others contain X-ray sources which are too luminous to be X-ray binaries and thus have to be AGN as well. Apart from the erroneously selected NGC 4472, all LINERs in our sample belong to this class of objects. Recent X-ray studies of local LINERs by Flohic et al. (2006) and Gonz´alez-Mart´ın et al. (2006) have found proof for the existence of an AGN in ∼ 60% of all observed objects. The remaining LINERs might contain Compton-thick AGN, but are probably powered by star formation. Sturm et al. (2006) have investigated the MIR properties of a sample of LINERs with Spitzer and found the population to be separated into IR-faint LINERs which are probably powered by an AGN and IR-bright LINERs which have an additional source of power. According to their analysis, the MIR SEDs of AGN powered LINERs are much bluer than the ones of IR-bright LINERs. Furthermore, the PAH flux ratios significantly differ. Interestingly, when looking at fig. 5.30, the VISIR photometry of NGC 4579 and NGC 7213, in both cases, shows a much bluer SED than the Spitzer IRS spectra. This could mean that with VISIR we see an AGN dominated LINER while Spitzer observes a contaminated object. The interpretation of the IRS spectra being contaminated would, however, require the absolute flux calibration of the spectrum of NGC 7213 to be wrong. This is not likely but possible as the mis-pointing in the case of NGC 4507 shows.
Even in LINERs that do contain AGN, Flohic et al. (2006) find that the ionising luminosities of these AGN, in many cases, are not sufficient to power the observed optical emission lines. They suggest that the line emission is either powered by radiation from hot stars or by a mechanical interaction with the AGN. The latter possibility, in turn, would agree nicely with the detailed analysis of NGC 4579 by Contini (2004).
Another LINER that has been extensively studied is NGC 1097. Mason et al. (2006) observed this AGN in the MIR with the Gemini South telescope. Their flux measurement
6.2 Implications for the state of the torus 83
exceed ours roughly by a factor of 2. This is probably due to the larger photometric aperture (300) they use. They compare their near- and mid-IR photometry to calculations based on the clumpy torus model by Nenkova et al. (2002) and come to the conclusion that the AGN is too weak to have a torus that could account for the observed IR emission. Please note that this result would still hold if our photometry was used. Mason et al. (2006) interpret the excess emission as originating in a starburst. Interestingly, they do not detect the 3.3 µm PAH feature in the nucleus even though it is detected in the larger aperture of Spitzer. This implies that either the UV radiation of the young stars destroys the PAH molecules close to the nucleus or that the MIR continuum in that location is not emitted by dust but by shock heated gas as in NGC 4579.
Our observations support the interpretation that LINERs are a mixed bag of differently powered sources. The most luminous LINER in our sample – NGC 7213 – lies very close to the correlation in fig. 5.32 and has the appearance of a regular Sy galaxy. The two less luminous objects NGC 1097 and NGC 4579, on the other hand, deviate significantly. As discussed above, their location in the LMIR−LXdiagram are consistent with the deviation
caused by contamination. If this contaminating emission is, at least in parts, caused by mechanical interaction with the AGN as indicated by the results of Contini (2004) and Flohic et al. (2006), the AGN in these low luminosity sources may be intrinsically different from the more luminous Seyfert type AGN.
7 Work on related projects
Knowledge is indivisible. When people grow wise in one direction, they are sure to make it easier for them- selves to grow wise in other directions as well. On the other hand, when they split up knowledge, concentrate on their field, and scorn and ignore other fields, they grow less wise – even in their own field.
Isaac Asimov: The Roving Mind (1983)
During the three years I have spent working on the mid-IR – hard X-ray correlation in AGN, I got involved in a couple of other projects as well. Those that are of importance in the context of this work will be briefly presented and discussed in this chapter.