The Astrophysical Journal, 660:117 Y 145, 2007 May 1 # 2007. The American Astronomical Society. All rights reserved. Printed in U.S.A. SPITZER OBSERVATIONS OF 3C QUASARS AND RADIO GALAXIES: MID-INFRARED PROPERTIES OF POWERFUL RADIO SOURCES K. Cleary,1 C. R. Lawrence,1 J. A. Marshall,2 L. Hao,2 and D. Meier1 Received 2006 April 28; accepted 2006 December 14 ABSTRACT We have measured mid-infrared radiation from an orientation-unbiased sample of 3CRR galaxies and quasars at 22:4 À1 redshifts 0:4 z 1:2 with the IRS and MIPS instruments on Spitzer. Powerful emission (L24 m > 10 WHz srÀ1) was detected from all but one of the sources. We fit the Spitzer data and other measurements from the literature with synchrotron and dust components. The IRS data provide powerful constraints on the fits. At 15 m, quasars are typically 4 times brighter than radio galaxies with the same isotropic radio power. Based on our fits, half of this difference can be attributed to the presence of nonthermal emission in the quasars but not the radio galaxies. The other half is consistent with dust absorption in the radio galaxies but not the quasars. Fitted optical depths are anticorrelated with core dominance, from which we infer an equatorial distribution of dust around the central engine. The median optical depth at 9.7 m for objects with core dominance factor R > 10À2 is 0.4; for objects with R 10À2,itis 1.1. We have thus addressed a long-standing question in the unification of FR II quasars and galaxies: quasars are more luminous in the mid-infrared than galaxies because of a combination of Doppler-boosted synchrotron emission in quasars and extinction in galaxies, both orientation-dependent effects. Subject headings: galaxies: active — galaxies: nuclei — quasars: general 1. INTRODUCTION Of all the identifying signatures of nuclear activity, only low- frequency radio emission originates far from the nucleus. Only From an observational perspective, active galactic nuclei (AGNs) low-frequency radio emission comes from outside the regions are diverse. The existence of relativistic jets in some AGNs, with where anisotropic emission and obscuration are potentially associated Doppler boosting (Blandford & Rees 1978), guaran- important. Low-frequency radio emission, therefore, pro ides a tees that some of this diversity of appearance must be due to ori- v unique way to obtain an orientation-unbiased sample of AGNs. entation (Readhead et al. 1978; Schever & Readhead 1979). It Only a small fraction of AGNs produce the prodigious radio is equally clear that not all differences can be due to orientation power of FR II galaxies and quasars. Nevertheless, the fact that and that there are real physical differences between AGNs. Much FR II galaxies and quasars can be studied in an orientation- observational and theoretical work over the last three decades has unbiased sample, selected on the basis of low-frequency radio been devoted to sorting out orientational and physical differences. The greatest successes have been achieved in the ‘‘unified model’’ emission, makes them uniquely valuable in separating the effects of orientation from physical differences. Visible light and X-ray of Fanaroff-Riley type II (FR II) galaxies and quasars (Orr & selected samples do not have this property. Browne 1982; Barthel 1989) and in the unification of Seyfert 1 and Seyfert 2 galaxies (see, e.g., Antonucci & Miller 1985). In The mid-infrared (MIR) and far-infrared (FIR) properties of these powerful radio sources are largely unknown. Their space both cases a dusty torus or disklike structure obscures optical and density is so low that only a few (e.g., 3C 405=Cygnus A) are at ultraviolet emission from the accretion disk and environs along low redshifts. The Spitzer Space Telescope (Werner et al. 2004) some lines of sight, accounting for apparent differences between promised a major advance in sensitivity over previous telescopes these classes of AGNs (see, e.g., Urry & Padovani 1995). and the capability to measure these objects. We therefore un- The FR II/quasar case is particularly important for the follow- dertook observations with Spitzer of powerful radio sources. The ing reason. These objects are selected to have low-frequency overall goal was straightforward: to measure for the first time the (178 MHz) radio luminosity greater than about 1025 WHzÀ1 srÀ1 (Fanaroff & Riley 1974). Low-frequency emission is synchro- MIR and FIR emission from an orientation-unbiased sample of tron emission from giant, optically thin clouds that therefore emit the powerful radio sources. isotropically. The low-frequency emission of dense and possibly More specific goals can also be considered. For example, the Doppler-boosted regions of these sources is suppressed by self- central proposition of FR II galaxy/quasar unification is that FR II absorption, which produces a 2.5 spectrum. As a result, bolo- galaxies are just quasars seen from an angle in which the optical- metrically insignificant but highly anisotropic emission, which to-UVemission of the central region is intercepted by the opaque, may dominate the source at higher frequencies, is almost always dusty torus, which reemits this radiation at longer wavelengths. insignificant at a few hundred megahertz or less. Moreover, high Conversely, quasars are just FR II galaxies seen closer to the jet spatial resolution observations of these sources with interfer- axis. Models suggest that the dusty torus becomes optically thin ometers allow one to measure and subtract the small contribution at wavelengths in the FIR (see, e.g., Pier & Krolik 1992; Granato of potentially anisotropic emitting regions at low frequency. & Danese 1994; Granato et al. 1997; Nenkova et al. 2002; Schartmann et al. 2005; Fritz et al. 2006), suggesting a direct test of FR II galaxy/quasar orientation-based unification. If radio lobe 1 Jet Propulsion Laboratory, California Institute of Technology, Pasadena, luminosity is an indicator (albeit time averaged) of central engine CA 91109; kieran.a.cleary@ jpl.nasa.gov. energy, then FR II galaxies and quasars with comparable isotropic 2 Astronomy Department, Cornell University, Ithaca, NY 14853. radio lobe luminosity should exhibit ‘‘comparable’’ isotropic FIR 117 118 CLEARY ET AL. Vol. 660 luminosity. Comparable does not necessarily mean equal, but ity and spectroscopic data provided by Spitzer provide additional there should be no FR II galaxies and quasars with low nuclear constraints in the infrared, allowing us to perform a spectral de- infrared luminosity at wavelengths where the emission is opti- composition of the thermal and nonthermal components of emis- cally thin. sion for the most powerful radio sources and also to estimate the This test has been performed by other investigators using in- silicate optical depth. frared measurements from the Infrared Astronomical Satellite In this paper we describe the selection methodology (x 2), (IRAS ) and the Infrared Space Observatory (ISO; Kessler et al. Spitzer observations (x 3), and data reduction (x 4) for our sam- 1996). Heckman et al. (1992) compared IRAS measurements of ple of powerful 3C radio sources. The photometric and spectro- 3CR quasars and radio galaxies (z > 0:3) selected on the basis of scopic results are presented (x 5.1) and the SEDs are fitted with a their 178 MHz radio flux density. They found that the rest-frame combination of synchrotron and dust components (x 5.2). We 6Y50 m emission is on average 4 times stronger in the quasars then explore what these results can tell us about the physical than in the radio galaxies, suggesting that either the quasars are conditions (x 6.1) and the origin of the infrared emission (xx 6.2 intrinsically stronger sources of MIR and FIR emission than qua- and 6.3) in these objects. Finally, the comparative infrared lu- sars or this emission is anisotropic. The latter requires large op- minosity of quasars and galaxies is discussed (x 6.4). tical depths in the FIR, possibly requiring implausibly large dust masses. The former possibility is problematic for pure orientation- 2. THE SAMPLE based unification of quasars and radio galaxies but may be due We require a sample selected at low frequency, with L178 MHz > to a significant nonthermal contribution in this wavelength 1026 WHzÀ1 srÀ1 and with a reasonable balance between FR II range from the radio jet. Hes et al. (1995) reanalyzed the IRAS 3 radio galaxies and quasars. 60 m data for 3CR quasars and radio galaxies in the redshift The sample of Barthel (1989) provides the ideal starting point. : < < : range 0 3 z 0 8. They confirm the finding of Heckman et al. It contains the 50 sources in the complete low-frequency 3CRR (1992) that 3CR quasars are systematically brighter than radio catalog of Laing et al. (1983) with 0:5 z 1:0. All have emit- galaxies in the MIR and FIR and conclude that a beamed 60 m ted radio luminosity L > 1026 WHzÀ1 srÀ1. Of the 50, 33 component may account for this difference. In subsequent work, 178 MHz are identified with galaxies and 17 with quasars, with classifica- Hoekstra et al. (1997) find that the observed IRAS 60 minfrared- tions based on the optical type designated by Laing et al. (1983). to-radio flux density ratios are consistent with an orientation- On the basis of radio morphology, all sources in the sample are based model incorporating beamed emission but conclude that FR II, as expected from the high radio luminosity. other processes such as torus optical depth also contribute. To reduce the observing time required, we reduced the sample The ISO photometry for four pairs of 3CR quasars and radio to 33 objects based on ecliptic latitude. Sources at high ecliptic galaxies matched in redshift and radio power was compared by latitude are easier to schedule with Spitzer.