Astron. Astrophys. 362, 75-96 (2000)

1. Introduction

Radio quiet and radio loud (not blazar) quasars (RQQ and RLQ, respectively) have similar spectral properties in the ultraviolet (UV), optical, and infrared (IR), but their radio powers differ by several orders of magnitude (Elvis et al. 1994). This divergence takes place at millimetre (mm) wavelengths. At these wavelengths the contribution from two emission components merge, namely the synchrotron emission dominant in the radio domain and thermal emission from cool dust (30-50 K) in the far-IR (Barvainis & Antonucci 1989). It is still not entirely clear whether the distinction between RLQ and RQQ is a consequence of differences in their central engines or whether it merely reflects differences in their environments. The primary observational distinctions in the IR domain, and the proposed physical mechanisms to explain them are studied here, using the new insights provided by Infrared Space Observatory ¹ (ISO; Kessler et al. 1996) measurements.

1.1. The radio emission

Two main types of RLQ can be distinguished on the basis of their radio spectrum: the flat spectrum radio loud quasars (FSRQ), and the steep spectrum radio loud quasars (SSRQ). FSRQ show highly-collimated structures and very compact features, with flat or inverted radio spectra. SSRQ have radio spectra dominated by synchrotron emission from extended radio lobes. The lobes and a radio core in the centre of these objects are signs of a relativistic jet. According to the unified scheme (Barthel 1989; Urry & Padovani 1995) FSRQ are the counterparts of SSRQ in which the jet is aimed at the observer.

The origin of the much weaker radio emission in RQQ is far less certain. The majority of the total radio emission from the RQQ comes from the compact features in the nucleus ( 1 kpc for unresolved regions, and at least 2 kpc for the resolved ones) rather than the body of the host galaxy (Kukula et al. 1998). It has been proposed that the activity in RQQ is supplied by a starburst, i.e. thermal bremsstrahlung and synchrotron emission coming from strongly radiative supernovae and supernovae remnants (SNRs) in a very dense environment where shocks accelerate electrons (Terlevich et al. 1992). Alternatively, if the energy supply arises from accretion onto a massive black hole, the radio emission from RQQ (as in RLQ) is caused by radio jets, but the bulk kinetic power of these jets is for some reason 10³ times lower than those of RLQ (Miller et al. 1993). This second hypothesis seems to be favored by recent studies, because of high brightness temperatures calculated (typical SN/SNRs have  K), by the evidence of a pc-scale jet (Blundell & Beasley 1998) and by observations of flat/inverted and variable radio spectra (Barvainis et al. 1996).

Recently, quasars with intermediate radio luminosities have been discovered and labeled Radio Intermediate Quasars (RIQ) (Francis et al. 1993, Falcke et al. 1995). RIQ may represent the Doppler boosted counterparts of radio quiet quasars. This hypothesis is suggested by the variability observed at radio wavelengths (Falcke et al. 1996).

1.2. The infrared emission

The presence of a dominant thermal (circumnuclear dust emission), or non-thermal (synchrotron radiation from the AGN) component in the IR continuum of quasars is still debated.

Many attempts to establish the origin of the IR emission in RLQ and RQQ have been performed through observations in the sub-millimetre (sub-mm) of quasars detected by IRAS (RLQ in Chini et al. 1989a, and Antonucci et al. 1990; RQQ in Chini et al. 1989b, Barvainis et al. 1992, Hughes et al. 1993, and Hughes et al. 1997; and both in Andreani et al. 1999, this last work is the only one based on an optically selected sample). The main test applied to recognize the presence of thermal emission in the IR spectra of the objects was based on the slope of the continuum emission () connecting the far-IR and sub-mm data. A steep, 2.5, continuum is strong evidence for thermal dust emission. Most of the sources studied have 2.5 and are consistent with a dominant self-absorbed synchrotron emission component. However, some RQQ have spectral slopes as steep as = 4.35, which, along with observations of strong molecular gas (CO) emission (Barvainis 1997), give strong support to a thermal mechanism as the origin of the far-IR component in RQQ.

Among the RLQ, is, at most, 0.9 for the FSRQ, and 1.1 for the SSRQ (Chini et al. 1989a). Variability, shape of the continuum spectral energy distribution and, in some cases polarization, indicate that the radio, mm and far-IR emission of FSRQ is dominated by the synchrotron process (Lawrence et al. 1991). On the contrary, many SSRQ show evidence of thermal emission: their far-IR spectra are brighter than extensions of the radio emission (Antonucci et al. 1990), suggesting a different origin than the non-thermal radio component; and the flux is constant, consistent with it arising from a region much larger than a light year (Edelson & Malkan 1987). Moreover, the spectral energy distributions of some RLQ, both FSRQ and SSRQ, show evidence for a galaxy component: reddening, residual starlight, molecular gas (Scoville et al. 1993), and some thermal dust emission in the near-IR (Barvainis 1987). Both components, non-thermal synchrotron radiation and thermal dust emission, are probably present at IR wavelengths, as observed in the RLQ 3C273 (Robson et al. 1986; Barvainis 1987).

1.3. Relation between the radio and infrared emission

A tight, linear correlation is observed between the far-IR flux and the radio fluxes in AGN (Sopp & Alexander 1991), suggesting a common origin. RQQ and RLQ occupy well defined regions in Log(L(IR))-Log(L(Radio)) space, and show a relation with a similar slope, just shifted to higher radio power by a factor 10⁴ in RLQ. RQQ show a similar relation as spirals, starbursts, and ultra luminous IR galaxies (ULIRG), suggesting that their IR emission may arise from sufficiently energetic star formation in the host galaxy (Sopp & Alexander 1991). However, the majority of the bolometric luminosity in over half of known ULIRG seems to arise from a buried AGN (Sanders 1999). Additionally, the variable and flat spectra, and high brightness temperatures shown by many RQQ at radio frequencies (Barvainis et al. 1996) suggest that the radio emission is related to the AGN rather than to a starburst.

1.4. Proposed scenarios

The unified model (Barthel 1989; Urry & Padovani 1995) predicts that similar disk-like dust distributions exist in both RQQ and RLQ. Orientation of the active nucleus, environment, and jet luminosity all affect the relative contributions of thermal and non-thermal sources to the observed infrared luminosity (Chini et al. 1989a).

Other scenarios have been proposed to explain the large differences in radio power between RQQ and RLQ: different spin of the central black hole (Wilson & Colbert 1995), or different morphological type of the host galaxy. Indeed, different radio powers are expected if one population of objects is fueled by mergers (ellipticals) and one is fueled by mostly internal processes within the galaxy (spirals) (Wilson & Colbert 1995). However, recent studies on the host galaxies of quasars indicate that the host galaxies of RQQ are in several cases elliptical and not always spiral galaxies (Taylor et al. 1996).

1.5. Open issues

A better knowledge of the radio and IR properties of quasars is required to test the unified model predictions, and answer the following questions:

1. What is the dominant mechanism emitting at IR wavelengths in RLQ and RQQ?

2. Do RLQ and RQQ have the same dust properties (temperature, source size, mass, and luminosity)?

3. Does an interplay between the radio and the IR components exist?

These questions can be addressed through the study of the spectral energy distributions (SED) of RLQ and RQQ. Here, we present the SEDs from radio to IR frequencies of a sample of 22 AGN (7 RQQ, 11 RLQ, 2 radio galaxies (RG) and 2 RIQ). The selected sample, even if incomplete and heterogeneous, is useful to address these questions thanks to several properties characterizing the sample (steep/flat radio spectra, radio loudness/quietness), and to the large amount of photometric data available in the radio, mm/sub-mm and IR domains. This work is based mainly on IR data provided by ISO. ISO data reduce the frequency gap between sub-mm and far-IR observations, better sample the IR spectral band with a larger number of filters than previous instruments, and increase the number of detected objects thanks to a higher sensitivity. The study of the IR emission of quasars will be extended in the future with the results of the European and of the U.S. ISO Key Quasar Programs providing a similar coverage of the IR SED for a larger sample of quasars (see first results in Haas et al. (1998), and Wilkes et al. (1999)).

© European Southern Observatory (ESO) 2000

Online publication: October 30, 19100
helpdesk.link@springer.de