arXiv:2011.03130v2 [astro-ph.HE] 11 Nov 2020 prtn nteANadi h otglx n their and radia- the galaxy disentangle host to the critical in is and mechanisms it a AGN inter-relationships, emission release the the al. in wave- both et understand of operating Brown they range better wide to As 2016; a over lengths, energy al. 2020). of 2012; et amount al. al. tremendous Bernhard 2012; role et et al. Shangguan Rosario the 2015; et 2012; 2019; Page of al. al. et see picture et Rovilos (e.g., Azadi 2012; coherent evolution al. no et this Aird yet in Dickinson as & play see Madau is AGN into and (e.g., review there formation decades turn past galaxy 2014), a the of can in significant for understanding been evolution dust our has (AGN; there in While and progress nucleus 2014). (SMBH), Best gas galactic & hole Heckman of active black accretion supermassive an by a hosts which galaxy ery ony atmr,M,220 USA 21250, MD, Baltimore, County, USA 85721, AZ, Tucson, Avenue, Cherry many lands M¨unchen, Germany b. Garching 85748 e tet abig,M,018 USA 02138, MA, Cambridge, Street, den rpittpstuigL using typeset Preprint ApJ to Submitted ti o ieyacpe httecne fev- of center the that accepted widely now is It ate Ashby Matthew 1 7 6 5 4 3 2 oea Azadi Mojegan etrfrAstrophysics for Center eateto hsc,Uiest fMrln Baltimore Maryland North of University 933 Physics, Arizona, of Department of University Observatory, Ger- Steward Bochum, Ruhr-University, Nether- The Institut, Groningen, Astronomisches of UK University Bristol, of Institute, University Kapteyn Laboratory, Physics Wills H.H. uoenSuhr bevtr,Karl-Schwarzschild-Str. Observatory, Southern European IETNLN H G N TRFRAINCNRBTOSTO CONTRIBUTIONS STAR-FORMATION AND AGN THE DISENTANGLING rsn h eut fAXE t otepnhoai Eso 20 of SEDs panchromatic the to emission fits 1 the ARXSED at for sample of accounts 3CRR results wavelengths t UV the the from for to present emission account ARXSE radio the to to for from wavelengths, accounts disk near-infrared component ARXSED radio homogeneous X-ray, At a and At optical-UV and lobes,jets). the medium hosts. clumpy (e.g., a co their structures combines multiple radio and state-of-t uses disent the AGN a ARXSED to from the introduce (ARXSED). essential we from is model Here emission it fitting wavelength. th distribution galaxies, of generating function energy host for a their as responsible and both mechanisms (AGN) nuclei emission galactic the constrain To ecieteSDo h ai-eetdANat AGN radio-selected the of SED the describe opc ai tutrs(oe o pt)aepeet efin ( We significantly contributes present. are structures spots) radio hot quasars’ (core, the structures radio compact h retto ftetrsaceinds osntln pwt the with up line not does Keywords: disk literatur the torus/accretion sample. in the our those of with agree orientation technique the fitting our from inferred ape matn h um-ae trfrainrt estimates at rate quasars formation radio-loud the star of submm-based SED the impacting sample, 1. 1 atnHaas Martin akBirkinshaw Mark , 3CRR 1 usr:RdoLu G:hg-esit–SD aais–galaxies – galaxies SED: – high-redshift AGN: – Radio-Loud Quasars: A eid Wilkes Belinda , INTRODUCTION T E tl mltajv 12/16/11 v. emulateapj style X | z < avr mtsna,6 Gar- 60 Smithsonian, & Harvard MSINO AI-ODQAASA 1 AT QUASARS RADIO-LOUD OF EMISSION . 5 S , efidta igepwrlwi nbet tterdoeiso when emission radio the fit to unable is power-law single a that find We . 2 lyHyman oley ´ 1 3 onaKuraszkiewicz Joanna , in Worrall Diana , > z n n httemda E fEvse l 19)i nbeto unable is (1994) al. et Elvis of SED median the that find and 1 6 aalMart Rafael , umte oApJ to Submitted ABSTRACT 2, > z 3 aah Abrams Natasha , .TeANtrsadaceinds parameters disk accretion and torus AGN The 1. bv h ufc fteds,icue naceindisk accretion in included disk, the of surface the above rmCmtnzto ftemlpoosb warm a by photons results it thermal al. that et of (k Petrucci possibility Comptonization the 2003; (e.g., investigate al. from excess we et this Czerny here there for 2006; 2018) While origin al. accepted 2009). et widely Crummy al. 2004; single, et extension Bianchi no Done the Gierli´nskiis 2005; & above (e.g., al. and et power-law over Piconcelli harder excess the soft of a show AGN neeg rdeti h msin ihhr X-rays hard distance with in emission, the results the to gradient in related ( the gradient temperature inversely of energy The are depth an SMBH. disk optical the the the from and in accre- temperature Rees gas 1976; The gaseous Sunyaev & a Shakura (e.g., 1984). 1973; emission by Thorne optical-X-ray & surrounded the PadovaniNovikov generates & is that Urry disk SMBH (e.g., tion galaxy the active an 1995), of picture wavelength. dard of function a as both from tion hshr -a msini sal ecie iha with ( energies described lower usually At 2001). is al. emission power-law. et simple X-ray Petrucci 1994; hard Maraschi & Zdziarski This Haardt & (e.g., Svensson tempera- disk inner 1993; (electron of the below hot up-scattering and X- a above Compton hard in (e.g., the The photons disk tures to disk 2018). outer due accretion Done the the is & in radiation Kubota 2012; arises ray al. radiation et UV Done and tical > msinfo h crto Disk: Accretion The from Emission T > 1 e ´ ınez-Galarza oahnMcDowell Jonathan , e)oiiaigi h ne einwieop- while region inner the in originating KeV) 2 0)t h um uioiyi afthe half in luminosity submm the to 70%) ∼ ∼ 0 e)adotclyti ooata situated that corona thin optically and KeV) 100 .– e)otclytik( thick optically keV) 0.1–1 httenntemleiso from emission non-thermal the that d eatANrdot--a spectral radio-to-X-ray AGN he-art a-nrrdwvlnts ARXSED wavelengths, far-infrared o iia ape.W n that find We samples. similar for e epeettemda intrinsic median the present We . eaceinds.A underlying An disk. accretion he 1 rmtehs aay eewe Here galaxy. host the from nlnto fterdojt in jets radio the of inclination ee Barthel Peter , 1 nrypwrn h active the powering energy e < Z < ienMeyer Eileen , nl h otiuin from contributions the angle ai-odqaasfo the from quasars radio-loud aito rmtetrs At torus. the from radiation pnnst elct the replicate to mponents consfrradiation for accounts D 2 1 afSiebenmorgen Ralf , H RADIO-X-RAY THE cie–AGN: – active : 7 4 ivniFazio Giovanni , τ ∼ ∼ e) many keV), 2 ntestan- the In 02)layer 10–25) 2 , 1 , 2 Azadi et al. models (e.g., Czerny et al. 2003; Kubota & Done 2018; the age of the electron population, our viewing angle, Petrucci et al. 2018). In these models, the UV and op- among many other factors. In a simplified picture the tical radiation originates from the outer regions of the radio spectrum of a lobe-dominated AGN can be de- α disk. The total radiative power of the accretion disk de- scribed with a power-law (Lν ∝ ν ). However, in AGN pends on the black hole mass, the mass accretion rate, with compact radio structures such as radio cores and the spin of the SMBH, radial dependence of the op- hot spots a superposition of various self-absorbed com- tical depth (e.g., Davis & Laor 2011; Done et al. 2012; ponents make the shape of the spectrum more complex. Kubota & Done 2018). Therefore, to model the radio emission multi-component Emission from The Torus: The SMBH and the ac- models with different spectral indices are required. The cretion disk are surrounded by an asymmetric dusty radiation from these structures eventually breaks at high structure (the torus) which absorbs some of the UV frequencies as a result of synchrotron radiation losses and and optical photons and re-radiates them at near- terminates at a cutoff frequency due to a drop in the infrared (NIR) to MIR-infrared (MIR) wavelengths (e.g., electron energy (e.g., Blandford & K¨onigl 1979; Konigl Neugebauer et al. 1979; Rieke & Lebofsky 1981). Ra- 1981). diation from the torus dominates a quasar’s emission What sparks the radio-loud phase in some AGN at 1–40 µm (e.g., Elvis et al. 1994; Netzer et al. 2007; is still unknown and may be stochastic by nature. Mullaney et al. 2011). The distribution of dust in the Some studies find that the triggering mechanism of torus has been the subject of many studies. Early the radio-loud phase is intrinsic to the AGN rather studies proposed a homogeneous structure in which than the host galaxies or the environments (e.g., the dust is smoothly distributed in a toroidal disk Kellermann et al. 2016; Coziol et al. 2017), while others (e.g., Pier & Krolik 1992; Efstathiou & Rowan-Robinson find cold star-forming gas in galaxies (e.g., Janssen et al. 1995; Fritz et al. 2006). However, these models are 2012; Best & Heckman 2012), the brightness of the clus- not able to describe some of the observed features, ter and density of the environment they live in (e.g., for example the 9.7 µm silicate absorption feature in Burns 1990; Best et al. 2007) increase the likelihood of Type 1 sources (e.g., Roche et al. 1991). A significant hosting a radio-loud AGN. The prevalence of radio-loud part of the MIR radiation comes from the polar re- AGN and the power of their radio emission (i.e., L1.4GHz) gions (e.g., Braatz et al. 1993; H¨onig et al. 2013) which correlates strongly with the intrinsic properties of the toroidal models could not explain. A clumpy circum- SMBH (e.g., SMBH mass, Best et al. 2005; Coziol et al. nuclear torus was then put forward as a possible solu- 2017). Radio-loud AGN, on average, have a higher Ed- tion (e.g., Nenkova et al. 2008; H¨onig & Kishimoto 2010; dington ratio (which is the ratio of the bolometric lumi- H¨onig et al. 2013). Recently Siebenmorgen et al. (2015) nosity to the Eddington luminosity) proposed a model in which the dust can be distributed in a homogeneous disk, or a clumpy medium, or a com- λEdd = Lbol/LEdd; LEdd ∝ MBH (1) bination of both (Siebenmorgen et al. 2005; Feltre et al. 2012). Models which combine both features success- bolometric luminosity and BH masses than their radio- fully reproduce the MIR spectra of AGN, including the quiet counterparts (e.g., Coziol et al. 2017). Such high 9.7. µm silicate absorption feature in Type 1 AGN, the accretion rates and radiative efficiencies occur relatively radiation from the hot dust close to the sublimation rarely, so the majority of the AGN population end up as temperature, and the MIR radiation from the ioniza- radio-quiet or weak radio emitters. tion cone (e.g., Braatz et al. 1993; Cameron et al. 1993; The multiple components described above emit radia- H¨onig et al. 2013). tion covering 10 decades of frequency from X-ray to ra- Emission from The Radio Structures: In addition to dio. To constrain these components we therefore need ob- the X-ray to IR emission noted above, a subset of AGN servations across a broad range of wavelengths. However, (∼ 15%, Kellermann et al. 1989; Urry & Padovani 1995) obtaining a multi-wavelength dataset is very challenging emit strongly at radio frequencies (for a recent review and in the case of quasars, variability adds more compli- see Hardcastle & Croston 2020) as a result of the in- cations. Over the past three decades, numerous studies teractions of relativistic electrons (or protons) with the have focused on spectral energy distribution (SED) anal- magnetic fields (see Ghisellini 2013, for a detailed re- ysis of AGN populations (e.g., Edelson & Malkan 1986; view). These radio-loud AGN launch powerful jets which Elvis et al. 1994; Richards et al. 2006; Shang et al. 2011; persist as highly collimated structures until they termi- Elvis et al. 2012; Hao et al. 2014). A number of these nate as bright shocks (hot spots) at the interface with studies attempt to understand whether quasars’ behav- the circumgalactic or intergalactic medium. Radio-loud ior can be described with an average SED. Elvis et al. AGN are classified on the basis of their radio power, (1994) presented the first high quality broad (X-ray to morphology, and radio spectral shape. They are desig- radio) atlas of quasar SEDs at z . 1 using then-current nated as FRII or FRI depending on whether their radio telescopes such as Einstein, the International Ultravio- 26.5 −1 −1 power P178MHzis in excess of 10 W Hz sr or not, let Explorer (IUE), and Infrared Astronomical Satellite respectively (Fanaroff & Riley 1974). FRII galaxies are IRAS. Elvis et al. (1994) presented median SEDs for the double-lobed sources and their jets are brightest at the radio-loud and radio-quiet quasars which have been ex- lobe edges. FRI galaxies’ jets are brightest in the cen- tensively used and overall works remarkably well at 0.1–1 ters of their lobes. The orientation of these AGN to the µm wavelength range for AGN of various luminosity and observer’s line of sight results in another classification as Eddington ratio (Elvis 2010). However, there is a large quasars and radio galaxies. dispersion around their median SED which can reflect The observed shape of the radio spectra varies with on the inferred properties of the quasars. Additionally, the size of the radio structure (length of the radio jets), their sample is not representative of the overall quasar Disentangling the AGN and star-formation contributions to the radio-X-ray emission 3 population and is favored towards X-ray bright, and blue piled the full X-ray to radio SED of 3CRR 8 (Laing et al. quasars (e.g., Jester 2005; Shang et al. 2011). 1983) quasars and radio galaxies at 1 redshifts concludes that orientation deter- for quasars and conclude that assuming an average SED mines the continuum shapes (e.g., Drouart et al. 2014; can potentially lead up to 50% errors in the bolomet- Podigachoski et al. 2015, 2016b). Here we expand on ric luminosities estimation. Hao et al. (2014) find that earlier SED fitting of the IR continuum to include the in X-ray selected, radio-quiet quasars the average SED X-ray to radio emission. In this paper we introduce de- has no significant dependence on redshift, bolometric lu- tailed, state-of-the-art AGN Radio-X-ray SED modeling minosity, SMBH mass, and the Eddington ratio, while (ARXSED) and focus on modeling the 3CRR quasars others find that local quasars templates may not be ap- and analysis of the physical properties of the AGN com- plicable to AGN at higher redshifts (e.g., Geach et al. ponents derived from our modeling (e.g., SMBH mass, 2011; Kirkpatrick et al. 2017). One of the many chal- Eddington ratio). SED fitting of the radio galaxies and lenges in AGN SED analysis is the contamination of the an investigation of the relation between the AGN and broadband photometry by host galaxy. Most of the stud- their host galaxies will be presented in future papers. ies noted here use scaling relations such as MBH − σ This paper is organized as follows: Section 2 describes (e.g., Hunt 2003) or Lhost − LAGN (Vanden Berk et al. the sample selection and the multi-wavelength data used 2006) or color-color diagnostics (e.g., Grewing et al. for the SED analysis. Section 3 presents the details of our 1968; Sandage 1971) to estimate the host galaxy con- SED modeling and the components used to describe the tamination. However, each of these scaling relations has emission from the AGN and the host galaxy at different large uncertainties that significantly effect the SED anal- wavelengths. Section 4.1 describes the fitting results for ysis. individual sources and results of the SED analysis are With the wealth of information stored in the photo- discussed in Section 5. We present a summary of our metric data obtained with modern telescopes with higher findings in Section 6. Throughout the paper we adopt a −1 −1 sensitivity and resolution , many studies have moved flat cosmology with ΩΛ =0.7 and H0=72 km s Mpc . towards fitting complex models to the broadband pho- tometry that can describe the radiation from the AGN 2. SAMPLE AND DATA COMPILATION and the host galaxy simultaneously and provide esti- The revised version of the 3CR catalog (3CRR) mates of the physical properties of each (i.e. SMBH (Laing et al. 1983) includes 180 FRII radio galaxies and mass, stellar mass). The galaxy SED models (e.g., MAG- quasars up to redshift z = 2.5 and is 100% complete to PHYS: da Cunha et al. 2008; FAST: Kriek et al. 2009; a 178 MHz flux density of 10 Jy. AGN have obscuration- CIGALE: Boquien et al. 2019) adopt libraries of galaxy dependent emission that results in strong selection effects templates with different stellar populations. These li- at most wavelengths (e.g., Azadi et al. 2017). However, braries are often built on assumptions on star forma- low-frequency (178 MHz) radio emission uniquely finds tion history (SFH), initial mass function (IMF), dust at- all the AGN being dominated by synchrotron emission tenuation law, etc which each have their own caveats from optically-thin radio lobes. In a sequence of papers, and uncertainties. In the case of active galaxies, ad- we focus on SED analysis of a subset of 3CRR AGN at ditional templates (i.e. torus, accretion disk) are re- 1 10Jy, DEC > 10◦ and Galactic latitude > 10◦ or < −10◦. 4 Azadi et al.

Table 1 Table 2 The 3CRR quasars modeled in this work Recent submm data obtained with the SMA or ALMA Name RA,DEC z Name ν (GHz) Flux (mJy) Telescope (J2000.0) 3C 2121 98 106 ± 5.3 ALMA 3C 009 00:20:25.2,+15:40:55 2.009 3C 2121 233 41 ± 2.1 ALMA 3C 014 00:36:06.5,+18:37:59 1.469 3C 2452 224 84 ± 1 SMA 3C 043 01:29:59.8,+23:38:20 1.459 3C 270.13 235 13.5 ± 0.3 SMA 3C 181 07:28:10.3,+14:37:36 1.382 3C 3184 315 2.3 ± 0.6 ALMA 3C 186 07:44:17.4,+37:53:17 1.067 3C 454.04 315 10.6 ± 1.4 ALMA 3C 190 08:01:33.5,+14:14:42 1.195 3C 191 08:04:47.9,+10:15:23 1.956 1) PI: Meyer; Proposal ID: 2019.1.01709.S 3C 204 08:37:44.9,+65:13:35 1.112 2) PI: Ashby; Proposal ID: 2019B-S034 3C 205 08:39:06.4,+57:54:17 1.534 3) PI: Ashby; Proposal ID: 2019A-S031 3C 208 08:53:08.8,+13:52:55 1.110 4) PI: Podigachoski; Proposal ID: 2015.1.00754.S 3C 212 08:58:41.5,+14:09:44 1.048 component used to model the emission from the host 3C 245 10:42:44.6,+12:03:31 1.029 galaxy. Additionally, we describe the methods used for 3C 268.4 12:09:13.6,+43:39:21 1.398 3C 270.1 12:20:33.9,+33:43:12 1.532 correcting for the obscuration from the torus, the host 3C 287 13:30:37.7,+25:09:11 1.055 galaxy, and the Milky Way absorption along the line of 3C 318 15:20:5.40,+20:16:06 1.574 sight. 3C 325 15:49:58.4,+62:41:22 1.135 4C 16.49 17:34:42.6,+16:00:31 1.880 3C 432 21:22:46.2,+17:04:38 1.785 3.1. 3C 454.0 22:51:34.7,+18:48:40 1.757 The AGN Components in ARXSED Here we describe the models accounting for the AGN ied, surprisingly many sources miss critical data needed emission as arising from three components: an accretion for estimating the basic parameters of their central en- disk, an obscuring torus, and radio lobes. gines and their host galaxies. For the SED analysis we have compiled the full X-ray to radio SED by combin- 3.1.1. The Accretion Disk Component ing archival data from Chandra, XMM-Newton, SDSS, For the accretion disk, we use the QSOSED model de- UKIRT, 2MASS, Spitzer, WISE, Herschel and multi- veloped by Kubota & Done (2018). The primary vari- frequency radio observations. The X-ray data used in able parameters in this model are presented in Table 3. this analysis is from Wilkes et al. (2013) and the opti- Below we describe the details of the model, the param- cal and UV data are gathered from the references avail- eters in Table 3 and our technique for constructing the able on NASA/IPAC Extragalactic Database (NED). accretion disk templates using this model. The NIR to FIR photometry used here is drawn from Kubota & Done (2018) assume that the emission from Podigachoski et al. (2015). Table 2 indicates the submm the accretion disk originates from three distinct regions: fluxes recently obtained from the SMA or ALMA for our The inner region, extending from the innermost stable analysis. Tables A1–A3 in appendix present the radio circular orbit (RISCO) to Rhot, the intermediate region observations (with their references) used in our analysis. extending from Rhot to Rwarm, the outer region from We assembled the final SEDs from these datasets by se- Rwarm to Rout. The Rwarm and Rhot define the bound- lecting the highest-quality photometry. The photometry ary of the hot and warm comptonization regions and Rout that significantly deviated from those within the same is the limit up to which the accretion disk extends (for wavelength range due to variability were removed. We more details see Appendix A of Kubota & Done 2018). excluded photometry with S/N< 3 at any wavelengths The inner region with a temperature of ∼ 100 KeV in- except at FIR and submm wavelengths, where the upper cludes the hot corona with no underlying disk (i.e., a limits are the only constraints available. Finally, we note truncated disk). The plasma in this region emits the that the UV-optical photometry is not contaminated by X-ray power-law component. In the intermediate re- the emission lines. We fitted a power-law continuum in gion the warm comptonization occurs and the soft X- IRAF (Tody 1986) to regions of the spectrum that were ray excess is produced. The electron temperature in uncontaminated by emission lines using continuum win- this region is ∼0.2 keV and and the optical depth is ∼ dows of Kuraszkiewicz et al. (2002). The power-law was 10−25 (e.g., Czerny et al. 2003; Gierli´nski & Done 2004; then binned into 10–15 continuum bins and included in Petrucci et al. 2013; Middei et al. 2018). The nature of our SEDs. the warm comptonization is not completely known (e.g., 3. Done et al. 2012) and it may be a result of the failed UV- SED COMPONENTS AND THE FITTING ROUTINE driven wind arising from the outer regions of the disk that One of the most important advantages of this study falls back down into the disk (Laor & Davis 2014). The relative to previous work is that it accounts for emission outer region of the disk is the standard optically thick over ten decades in frequency, from X-ray to radio accretion disk dominated by the blackbody emission. wavelengths. Because both the AGN and host galaxies The QSOSED model fixes the accretion disk pa- contribute throughout the electromagnetic spectrum, rameters to typical AGN values of kTe,hot = 100 keV, it is critical to disentangle the contributions from each kTe,warm = 0.2 keV, Γwarm = 2.5, Rwarm = 2Rhot , as a function of wavelength in order to quantify the Rout = Rsg which are respectively, the electron tem- contributions of the various physical mechanisms to the perature for the hot and warm comptonization compo- overall energy budget. In this section, we first describe nents and the spectral index of the warm comptoniza- the components used in our SED model (ARXSED) to tion component. The spectral index of the hot comp- account for the AGN emission, and then describe the tonization component, Γhot, is calculated via equation Disentangling the AGN and star-formation contributions to the radio-X-ray emission 5

rest-frame rest-frame 10µm 1µm 0.1µm 10nm 1keV 10keV 100keV 10µm 1µm 0.1µm 10nm 1keV 10keV 100keV 47 47 9 M = 5x10 MO • log (λ ) = 0.39 BH 9 Edd M = 1x10 MO • log (λ ) = 0.0 BH 8 Edd M = 5x10 MO • log (λ ) = -0.5 BH 8 Edd M = 1x10 MO • log (λ ) = -1.0 46 BH 7 46 Edd M = 5x10 MO • log (λ ) = -1.5 BH 7 λEdd MBH = 1x10 MO • log ( Edd) = -1.65

45 45 [erg/s]) [erg/s]) ν ν L L ν 44 ν 44 log ( log (

43 43

42 42 13 14 15 16 17 18 19 20 13 14 15 16 17 18 19 20 log(ν [Hz]) rest-frame log(ν [Hz]) rest-frame rest-frame rest-frame 10µm 1µm 0.1µm 10nm 1keV 10keV 100keV 10µm 1µm 0.1µm 10nm 1keV 10keV 100keV 46 46 spin = 1.0 cosi = 1.00 spin = 0.9 cosi = 0.80 spin = 0.8 cosi = 0.60 spin = 0.7 cosi = 0.40 spin = 0.5 45 cosi = 0.20 cosi = 0.05 45 spin = 0.0 [erg/s]) [erg/s])

ν ν 44 L L ν ν 44 log ( log ( 43

43 42 13 14 15 16 17 18 19 20 13 14 15 16 17 18 19 20 log(ν [Hz]) rest-frame log(ν [Hz]) rest-frame Figure 1. The effect of variation of each of the primary variable parameters in the QSOSED model (Kubota & Done 2018) on the accretion disk SED while other parameters are kept constant. Top-Left: An increase in the BH mass results in a more luminous and cooler accretion disk. Top-Right: An increase in the mass accretion rate results in a more luminous and hotter accretion disk. Bottom-Left: An increase in the spin results in a more luminous and hotter accretion dis with higher radiative efficiency. Bottom-Right: An increase in the inclination angle from face-on to edge-on results in a more luminous optical-UV SED.

Table 3 The initial estimates, range of variation and steps used in constructing the accretion disk templates as well as the acceptable range for each parameter in the QSOSED model Parameter Initial estimate Range of variation Steps Acceptable range

log(MBH ) M0 (C IV or Mg II)[log(M0)-1,log(M0)+1] 0.2 dex [7,10] log(λEdd)–[−1.65, 0.39] 0.2 dex [-1.65,0.39] spin (a) – [0.75, 0.998] 0.02 [0,0.998] inclination angle (θD) θT [θT -12,θT +12] θT dependent [0,90] 6 in Kubota & Done (2018). As described above, the of the accretion disk and on its SED are intertwined. The Rwarm and Rhot define the boundary of the hot and top-left panel illustrates how an increase in the BH mass, warm comptonization regions (for more details see Ap- while logm ˙ , and other parameters are constant results in pendix A of Kubota & Done 2018) and Rsg is the self– a more luminous and cooler accretion disk. The top-right gravity radius (see equation 4 of King 2016). The panel indicates that an increase in the Eddington ratio QSOSED model includes reprocessed radiation which is results in a more luminous and hotter accretion disk. the fraction of the hot comptonization component that il- Increasing the spin (bottom-left) can impact the geom- luminates the warm comptonized material and cool outer etry by moving the RISCO closer to the SMBH and the disk (which has a temperature of a few thousands of shape of the SED by moving the peak towards higher en- kelvin). ergies (similar to the Eddington ratio) and increase the The primary variable parameters in the QSOSED radiative efficiency. The bottom-right panel of Figure il- model are the suppermassive black hole mass MBH , the lustrates, while all other parameters are kept constant, a rate of accretion (logm ˙ , which is the Eddington ratio face-on observer sees more of the UV bump compared to in log space, see equation 1), the inclination angle and an edge-on observer. ≡ 2 the dimensionless spin parameter a Jc/GMBH where To fit the optical-Xray SED of each quasar, we con- J is the angular momentum of the BH. Figure 1 illus- struct a set of QSOSED templates in XSPEC9 covering trates the effect that variation of each parameter has on a range for each parameter and using prior information the shape of the optical, UV and X-ray synthetic spectra 9 XSPEC is an interactive program used to fit models to energy while keeping the other parameters constant. Overall, spectra from optical to gamma-rays. For more information see the effect of varying these parameters on the geometry https://heasarc.gsfc.nasa.gov/xanadu/xspec/ 6 Azadi et al.

rest−frame rest−frame 100µm 10µm 1µm 100µm 10µm 1µm

° 15 ° 15 τ τ θ τ τ θ VC = 38.5%, C = 13.5, D = 100, T = 43 rin = 772 x 10 cm, C = 13.5, D = 100, T = 43 44 rin = 1545 x 10 cm 44 VC = 77.7 % 15 rin = 1000 x 10 cm VC = 38.5 % 15 rin = 772 x 10 cm VC = 7.70 % 15 rin = 514 x 10 cm VC = 1.50 % 15 rin = 300 x 10 cm 43 43 [erg/s]) [erg/s]) ν ν L L ν ν log ( log (

42 42

12 13 14 12 13 14 log(ν rest−frame [Hz]) log(ν rest−frame [Hz])

rest−frame rest−frame 100µm 10µm 1µm 100µm 10µm 1µm

15 τ θ ° 15 τ θ ° τ rin = 772 x 10 cm, VC = 38.5%, D = 100, T = 43 τ rin = 772 x 10 cm, Vcloud = 38.5%, C = 13.5, T = 43 44 C = 45.0 44 D = 1000 τ τ C = 13.5 D = 300 τ τ C = 4.50 D = 100 τ τ C = 0.00 D = 30 τ D = 0 43 43 [erg/s]) [erg/s]) ν ν L L ν ν log ( log (

42 42

12 13 14 12 13 14 log(ν rest−frame [Hz]) log(ν rest−frame [Hz])

rest−frame 100µm 10µm 1µm

° ° 15 τ θ θ rin = 772 x 10 cm, VC = 38.5%, C = 13.5, T = 43 44 T= 86 θ ° T= 73 θ ° T= 60 θ ° T= 43 θ ° T= 19 43 [erg/s]) ν L ν log (

42

12 13 14 log(ν rest−frame [Hz])

Figure 2. The impact of variation of one parameter on the torus SED while keeping the other parameters constant. Top-Left: As Rin increases the torus moves away from the central engine, thus the emission shortward of 10 µm decreases, and the peak of the torus emission moves to higher wavelengths. Top-Right, Middle-Left: As VC and/or τC increase the higher density of clumps at radii close to the central engine results in more absorption so the silicate feature changes from emission to absorption. Middle-Right: An increase in τD results in more scattering from the disk surface at all radii, therefore results in more emission shortward and longward of 10 µm. Bottom: A change in the inclination angle from face-on to edge-on results in less pronounced emission shortward of 10 µm and change of the silicate emission feature to absorption. when it is available as starting points for the templates. Barthel et al. 1990) which we use to estimate the BH To construct the templates for each quasar we vary each masses. For the remaining quasars without spectro- parameter with a fixed step within their acceptable range scopic information, we use the median BH mass of the in QSOSED. Table 3 summarizes the initial estimates, 12 sources with C IV or Mg II lines as the initial mass range of variation and steps used in our procedure and estimate. We note that BH mass measurements based the acceptable range for each parameter in the model. on the high ionization C IV line have larger errors than Out of 20 sources in our sample, 6 (3C 9, 3C 191, those estimated from the Mg II or Balmer lines (e.g., 3C 205, 3C 270.1, 3C 432, 3C 454.0) have broad Sulentic et al. 2002; Baskin & Laor 2005; Shen 2013). C IV λ1548 line and 6 (3C 14, 3C 181, 3C 186, 3C 204, However, we only use this mass as initial estimate for 3C 245, 3C 268.4) have Mg II λ2800 line available making the accretion disk templates. We allow MBH to (obtained from SDSS archive, private commutation or vary within ± 1 dex around the initial mass estimates Disentangling the AGN and star-formation contributions to the radio-X-ray emission 7

Table 4 The free parameters and their acceptable values in Siebenmorgen et al. (2015) torus library Parameter Acceptable values 1 15 Inner radius ( rin) 300, 514, 772, 1000, 1545 (×10 cm) Cloud volume filling factor (Vc) 1.5, 7.7, 38.5, 77.7 (%) Cloud optical depth (τC ) 0, 4.5, 13.5, 45 Disk optical depth (τD) 0, 30, 100, 300, 1000 ◦ inclination angle (θT ) 19, 33, 43, 52, 60, 67, 73, 80, 86 ( ) 2 1. The inner radius is the dust sublimation radius and scales with the total AGN luminosity as Rin ∝ LAGN . The values listed here 11 are for an AGN of a luminosity of 10 L⊙ in steps of 0.2 dex. Since there is no prior informa- the torus moves farther from the central engine (top- tion on logm ˙ we allow it to vary over the entire accept- left panel) therefore the emission hotter dust grains be- able range in QSOSED in steps of 0.2 dex. The quasars come less pronounced and the peak of the torus SED in our sources are all radio-loud thus are expected to moves to longer wavelengths. The impact of variation of have high spin values (see Reynolds 2019, and references the filling factor and optical depth of the clumps (top- therein). Therefore we use a limited range from 0.75 to right and middle-left) is intertwined. An increase in VC 0.998 in steps of 0.02 for spin. For the inclination angle, and/or τC while the inclination angle and Rin are con- we use the best-fit value from the torus model (Section stant changes the SED from one with a silicate emission 3.1.2) then construct the templates by varying the incli- feature at 9.7 µm to a one with an absorption feature. ◦ nation angle within ±12 of the initial estimate. How- In fact an increase in VC and/or τC when the homo- ever, rather than using a fixed step size we determine geneous disk is unchanged result in more absorption at 7 values within ±12◦ of the initial estimate empirically. shorter wavelengths ( because of a higher density of the Overall, with this approach, we construct ∼ 11000 tem- clouds closer to the central engine) and the peak of the plates to fit the optical-Xray SED of each quasar. torus emission moves to higher wavelengths. An increase in the opacity of the homogeneous disk (τD) results in 3.1.2. The Torus Component more scattering from the disk surface which acts simi- Here we adopt the two-phase AGN torus model devel- larly to adding more hot and cold dust grains (for a face- oped by Siebenmorgen et al. (2015) which assumes the on observer). Therefore the emission from short and long dust around the AGN can be described as a clumpy wavelengths both increases. We note that this will not medium or a homogeneous disk, or as a combination of be the same for an edge-on observer (see Figures 4 and 5 both. Table 4 lists the primary variable parameters and in Siebenmorgen et al. 2015) As the observer’s inclina- their acceptable values in the model. Below we describe tion angle changes from face-on towards edge-on (bottom the details of the model, the variable parameters. panel), the observer sees less of the emission from the In Siebenmorgen et al. (2015) the inner radius (Rin) of inner parts of the torus (short wavelengths) and rather the dusty structure is defined by the sublimation temper- than a silicate emission feature sees an absorption fea- ature of the dust grains. The outer radius of the structure ture. We note in our fitting procedure in general we do is chosen to be Rout = Rin × 170 which is large enough not have any prior information to limit the torus parame- to have negligible impact on the observed features in ters but see Section 4.1 for a few cases where the observed the SED. The density of the dust decreases with radius. silicate feature helped to constrain the torus model. The dust grains are fluffy mixtures of silicate and amor- As noted in the Introduction the MIR emission from phous carbon grains (Kr¨ugel & Siebenmorgen 1994) in- the torus comes from the absorbed and reprocessed UV stead of the interstellar medium (ISM) dust grains (e.g., and optical photons radiated by accretion disk. There- Siebenmorgen et al. 2014). These fluffy grains are more fore, the SED shape of the primary source of emission efficient aborbers and have higher submm emissivities might influence the reprocessed MIR radiation. The pri- than the diffuse ISM grains (for more details see Section mary source of radiation in Siebenmorgen et al. (2015) 2.6 in Siebenmorgen et al. 2015). is the Rowan-Robinson (1995) model, which describes The free parameters of the torus model, Rin, the op- the optical-UV continua of quasars with a broken power- tical depth of the homogeneous disk mid-plane (τD), the law function. As noted in Section 3.1.1 we use the optical depth of the clumps (τC ), the volume filling factor Kubota & Done (2018) accretion disk model. We exam- (VC ) and the inclination angle (θT ) and their accepted ined the impact of various accretion disk models (includ- values are listed in Table 4. Varying these parameters ing a simple blackbody, Kubota & Done 2018 and a few within their acceptable range makes a library with 3600 others) on the torus SED, and found that the SED shape templates. We note that these templates are built for longward of 1 µm (rest-frame) is independent of the cho- 11 an AGN of luminosity 10 L⊙. However, the SEDs are sen accretion disk template at wavelengths shorter than scale-invariant so the actual inner radius scales with lu- 1 µm. minosity of the primary source (i.e., the accretion disk) (e.g., Suganuma et al. 2006; Kishimoto et al. 2011): 3.1.3. The Radio Component R ∝ L2 (2) In this section we describe the four models developed in AGN in our study to model the radiation from quasars’ radio Figure 2 illustrates the effect that variation of each structures. To constrain the radio fits parameters and to the parameter on the shape of the NIR-FIR SED, while understand which of the four models better describe the keeping the other parameters constant. As Rin increases radio emission we use the radio properties listed in Ta- 8 Azadi et al.

Table 5 Radio properties of the quasars in our sample 1 2 3 4 5 6 Name Type RCD Projected jets length Inclination angle De-projected jets length Range of de-projected jets length (kpc) (◦) (kpc) (kpc) 3C 009 - 0.009 85 52 ±10 151 [135,178] 3C 014 – 0.018 213 42±10 308 [261,389] 3C 043 CSS < 0.060 26 > 26 < 49 – 3C 181 – 0.009 59 49±10 65 [57,77] 3C 186 CSS 0.042 16 42±10 20 [17,25] 3C 190 CSS 0.098 25 27±10 124 [94,194] 3C 191 – 0.102 34 31±10 81 [64,116] 3C 204 – 0.087 296 34±10 543 [437,748] 3C 205 – 0.0309 154 38±10 251 [208,330] 3C 208 – 0.105 107 32±10 172 [136,243] 3C 212 – 0.204 73 20 ±10 216 [147,425] +10 3C 245 – 1.950 40 4 −4 1067 [307,–] 3C 268.4 – 0.091 85 27±10 205 [155,319] 3C 270.1 – 0.282 103 16±10 374 [235,987] 3C 287 CSS < 0.1027 8– – – 3C 318 CSS < 0.138 9 58 ±10 8 [7,9] 3C 325 – 0.003 124 – – – 4C 16.49 – 0.052 145 39±10 216 [180,281] 3C 432 – 0.025 111 46±10 155 [134,190] 3C 454.0 CSS < 0.339 10 > 14 < 46 – 1) Compact Steep Spectrum (Wilkes et al. 2013) 2) Radio core dominance (see equation 3) 3) Projected radio sizes, measured lobe-to-lobe from high resolution images at 5Hz 4) Jets inclination angle (Marin & Antonucci 2016) 5) De-projected radio sizes, measured lobe-to-lobe at 178 MHz 6) The range of the de-projected sizes estimated from the errors on the inclination angles 7) We did not find the RCD estimate for 3C 287 in the literature, so we adopted the average values of the other CSS quasars as an upper limit. ble 5. The first column of the table gives the source name, an optically thin region with the radiative transfer equa- second column indicates whether the quasar is classified tion (for more details see Section 4.4–4.7 in Ghisellini as a compact steep spectrum (CSS) source, third column 2013): lists the radio core dominance parameter (Orr & Browne

1982), i.e., the ratio of radio-core to extended radio-lobe ǫ(ν) −τν I(ν)= (1 − e ); τν ≡ Rκν (5) emission, both measured at 5 GHz: κν R = L (5 GHz)/L (5 GHz) (3) CD core lobe in which τν is the spectral optical depth, R is the size from (Wilkes et al. 2013). The fourth column gives of the emitting region. ǫ(ν) is the emissivity, and κν is the projected radio jet length measured lobe-to-lobe at the specific absorption coefficient and are described as 178 MHz taken from the compilation of C. Willott 10 . The fifth column lists the radio jets inclination angle, ǫ(ν) ∝ B(p+1)/2 ν−(p−1)/2 (6) accurate to within ±10◦ from Marin & Antonucci (2016) (estimated from the radio core fractions – see their Sec- (p+2)/2 −(p+4)/2 tion 3). The de-projected radio jet lengths (listed in sixth κν ∝ B ν . (7) column) were calculated using the projected jet sizes and dividing them by the sine of the jets inclination angle. In a self-absorbed regime τ ≫ 1, therefore the second The last column indicates the range of the de-projected ν component in the radiative transfer equation becomes jets length taking into account the errors on the jet in- negligible and equation 5 can be simplified (see Section clination angle. 4.5 in Ghisellini 2013) to: To fit the radio emission from the quasars in our sample we start by considering a relativistic electron population ∝ −1/2 5/2 with a power-law energy distribution in a magnetic field I(ν)= I0(ν) B ν (8) of B: Setting ντ to be the frequency at which the a transition from optically thick to thin occurs, then N(E) ∝ E−p (4) τ ≡ Rκ = 1 (9) Synchrotron emission is generated by these relativis- νt νt tic electrons spiraling around magnetic field lines. Al- and using equation 7 we then obtain though sychrotron radiation is absorbed by the electrons in optically thick regions (synchrotron self-absorption), ν ∝ [RB(p+2)/2]2/(p+4) (10) radio photons from optically thin regions reach the ob- t server. We can formulate synchrotron emission transi- therefore tioning from an optically thick, self-absorbed region to

10 ν ν http://astroherzberg.org/people/chris-willott/research/3crr/ τ ≡ Rκ = ( t )(p+4)/2 = ( t )(p−1)/2+5/2 (11) ν ν ν ν Disentangling the AGN and star-formation contributions to the radio-X-ray emission 9

Now if we consider α2 = −(p − 1)/2 as the spectral Model 4–A Parabola with a Cutoff & A Double Power- index in the optically thin region and α1 = 5/2 as the Law with an Exponential Cutoff: In quasars with spectral index in the optically thick region we can rewrite bright cores, the superposed self-absorbed components equation 5 as may generate spectral shapes not amenable to fitting with these simple models. Therefore, in our fourth model ν νt I(ν) ∝ ( )α1 [1 − exp(−( )α1−α2 )] (12) we consider a combination of Eqs. 13 and 15, where νt ν Eq. 13 is solely used to describe the radio core emission, However, as the electron population ages it loses its and Eq. 15 describes radiation from the lobes/hot spots. energy and eventually the synchrotron radiation termi- We note that in this case, the core and lobe components nates at a cutoff frequency. Therefore an additional term each have their own cutoff frequency. We fit these four models to the observed radio pho- is required to account for this energy loss (Polletta et al. MPFIT 2000): tometry using the nonlinear least-square fitting function in IDL (Markwardt 2009). The spectral slopes α1 and α2 in each model as well as β and νt are free pa- ν − ν ν α1 t α1−α2 ν 10 Lν ∝ ( ) [1 − exp(−( ) )] e cutoff (13) rameters, and the cutoff frequency is limited to 10 < 14 νt ν νcutoff < 10 Hz. Equation 13 describes radiation from both optically thick and thin regions which terminates at high ener- 3.2. The Host Galaxy Component gies due to energy loss. This equation is based on many In this study we use the MAGPHYS SED code assumptions including, the charged particle is an elec- (da Cunha et al. 2008, 2015) to account for the contribu- tron, the electron population’s energy has a power-law tion from the host galaxy. MAGPHYS is a widely used distribution, the minimum energy condition between the galaxy SED code capable of accounting simultaneously relativistic particles and the magnetic field is met, the for different levels of star formation activity, stellar pop- source is homogeneous and therefore α =2.5. In order 1 ulations, dust obscuration, and star formation histories to describe the radio emission from the quasars in our for galaxies at different redshifts. We did not use MAG- sample, we use the general form presented in equation PHYS to model AGN emission in our sources, because 13 and modify it based on different conditions. Model 1–A Single Power-Law with an Exponential we handled that as described in Section 3.1. MAGPHYS is built upon the energy balance tech- Cutoff: In our first model we consider an AGN in which nique, which links the UV and optical emission from the radio emission is dominated by radiation from the op- the young stellar population to the IR emission from tically thin lobes. In this case, a power-law can success- dust. In other words, starlight is the only source of fully describe the radio emission (Polletta et al. 2000): dust heating and the energy absorbed by dust is equal ν α − ν to the re-radiated energy, and therefore no energy trans- L ∝ ( ) 2 e νcutoff (14) ν ν fer is involved. The stellar emission from UV to NIR cutoff wavelengths is produced using the Bruzual & Charlot Model 2–A Double Power-Law with an Exponential (2003) spectral population synthesis model (assuming Cutoff: The added presence of compact structures such a Chabrier 2003, IMF) attenuated by dust following as radio cores and hot spots may cause the spectral shape Charlot & Fall (2000). The model assumes the young to deviate from a single power law. Therefore, in our stars form in dense clouds (i.e., giant molecular clouds); second model, we assume a double power-law similar to when younger their emission is attenuated by the dust in Eq. 13. However, in our model, α1 is a free parameter their birth cloud and the ambient ISM, but as they age and is not fixed at 2.5 (since we do not expect a sin- the birth clouds disappear on a time-scale of 107 yr, and gle, homogeneous component) and νt is the transition the stellar emission is then absorbed only by the diffuse frequency from α1 to α2 dominated regions. As we dis- ISM (Charlot & Fall 2000). MAGPHYS considers four cuss in Section 4.1 indeed some of the sources in our dust components, each including grains with a different sample with bright hot spots can not be fitted with a size and temperature: polycyclic aromatic hydrocarbons single power-law and require a second component with a (PAHs) grains, hot grains with temperatures in the range different spectral slope. 130–250K, grains in thermal equilibrium with tempera- Model 3–A Parabola with a Cutoff: Relaxing the ho- ture of 30–60K, and cold dust grains with adjustable mogeneous assumption, in our third model instead of a equilibrium temperature in the range 15 –25 K. The stel- broken power-law we assume a parabolic function which lar birth clouds contain the first three dust components approximates a superposition of multiple power-law com- and cold dust only exists in the ambient ISM. For more ponents some optically thick and some thin (Cleary et al. details, refer to da Cunha et al. (2008). 2007): Although the original da Cunha et al. (2008) MAG- PHYS code covers a broad wavelength range from 912 A˚ − ν 2 ν to 1 mm, here we use the updated da Cunha et al. (2015) log L ∝−β(log ν − log ν ) + log(e cutoff ) (15) ν t version, which includes the galaxy contribution at ra- β in this equation indicates the curvature of the dio wavelengths (for more details see Section 3.2 in parabola. da Cunha et al. 2015). In addition to broader wave- Overall, Eqs. 13 or 15 can successfully describe the length coverage, da Cunha et al. (2015) includes a con- emission from a source with a mix of optically thick and tinuous delayed exponential SFH: thin regions (e.g., a lobe-dominated quasar with bright hot spots or a moderately bright core). Ψ(t) ∝ γ2t exp(−γt) (16) 10 Azadi et al. in which t is the time since the onset of star forma- In this Section we describe our fitting methodology and tion and γ = 1/τSF is the inverse of the star formation the steps in which the various attenuation and absorption time scale (see Lee et al. 2010; da Cunha et al. 2015). correction from Section 3.4 are applied. The X-ray lumi- da Cunha et al. (2015) also include absorption by the in- nosities used in the fitting analysis are corrected for both tergalactic medium (IGM) which can strongly impact the intrinsic and the Milky Way absorption (Wilkes et al. UV emission in high-redshift galaxies such as the ones 2013). We correct the X-ray emission in advance for treated here. the gas absorption since we have no independent way to Although the da Cunha et al. (2015) model includes estimate this beyond the X-ray data. Additionally, the galaxy emission at radio wavelengths, the radio emission photometry at UV–NIR wavelengths are corrected for (both thermal and non-thermal components) in galaxies the Milky Way absorption (see Section 3.3). is ∼ 4–5 orders of magnitude lower than that in radio- Given that our sample consists of radio-loud quasars, loud AGN. Therefore AGN are the dominant sources of AGN emission dominates that of the host galaxy over radio emission in the present analysis. most of the SED. Therefore, we allow the contribution of the AGN to the photometry to vary from 95% to 65% 3.3. Dereddening and Absorption Corrections (in bins of 5%) at submm to UV wavelengths (at radio and X-ray wavelengths the host galaxy contribution is In order to obtain the intrinsic SED of the AGN in our sample, it is necessary to correct the photometry for orders of magnitude smaller than the AGN). The AGN the reddening and absorption occurring at various wave- components dominating at submm to UV wavelengths , torus and accretion disk, are allowed to vary indepen- lengths. In brief, these corrections include, correcting the X-ray observations for the host galaxy and the Milky dently from each other. The range of normalization of Way absorption, correcting the UV–NIR photometry for the contribution of the AGN components results in an iterative process in which the steps below are followed: the Milky way absorption, and correcting the optical-UV Step 1: radiation from the accretion disk for the dusty torus ab- We start the fitting procedure by determin- sorption. ing the template from Siebenmorgen et al. (2015) which best fits the photometry associated with the AGN at X-ray photons are absorbed by the gas in their host galaxies (i.e., intrinsic) and along the line of sight in the NIR–FIR wavelengths. Siebenmorgen et al.’s templates are normalized to a primary AGN of the luminosity of Milky Way. In this study we adopt the best estimates of 11 10 L⊙ (Section 3.1.2). Therefore to fit our quasars, the intrinsic X-ray luminosity from Wilkes et al. (2013), i.e., corrected for both intrinsic and Galactic absorp- we integrate the AGN luminosity within the range of 2– tion. Additionally, we correct the UV–NIR (0.91 - 13.0 45 µm in rest-frame (at which wavelengths the torus is the dominant radiation source) and normalize the torus µm) photometry for absorption by the Milky Way using ∝ −0.7 templates according to this integrated luminosity. We the attenuation law of τλ λ from Charlot & Fall determine the best-fit torus using χ2 minimization. (2000). Step 2: We correct the optical-UV photometry asso- To estimate the extinction of the accretion disk emis- ciated with the accretion disk for the torus reddening sion from the obscuring structure in the torus we use the implementing the correction factor obtained from equa- method of Siebenmorgen et al. (2015, see also Kr¨ugel tion 17. To do this we assume the line-of-sight to the 2009). According to this method the effective optical accretion disk is the same as that for the torus. Con- depth for any templates (with any combination of dust sidering that AGN are known to vary and the fact that clouds and homogeneous disk) can be obtained via a com- some of sources are missing UV photometry, the accre- parison to the same template in the absence of dust (see tion disks templates are normalized to the average flux Section 2.7 in Siebenmorgen et al. 2015): in the 0.2–0.9 µm (rest-frame) band. We identify the best-fit accretion disk template via χ2 minimization. Fluxtemplate τeff = − ln (17) Step 3: We fit the radio photometry with the four mod- Fluxnodust els described in Section 3.1.3, and determine the best-fit 2 Thus, we first determine the template from using χ minimization. We normalize the radio models Siebenmorgen et al. (2015) which best fits the NIR–FIR to the 5 GHz photometry. In the fourth radio model, we data corresponding to the AGN emission (not the host use the 5 GHz core flux to normalize the core component galaxy, see Section 3.4 for more details) and then apply and the remaining flux to normalize the parabola com- equation 17 to estimate the effective optical depth. ponent. For the radio models, we inspected the radio The UV and optical photometry (corresponding to the images in several bands (178 MHz, 5 GHz, 8 GHz, and accretion disk emission) is then corrected by a factor 15 GHz, if available) to reconcile multi-component mod- of eτeff for the absorption in the torus. We note that, els with different spectral indices are required with the depending on the best-fit, torus template τeff may be presence of bright hot spots and/or cores emission. a negative or positive number indicating scattering or Step 4: We subtract the best-fit AGN component (ra- absorption by the dust structure. An edge-on observer dio, torus and accretion disk) from the total observed is not able to see the scattered light from the torus and photometry and fit this residual with MAGPHYS to ac- is mainly affected by absorption, i.e., a positive τeff , count for the host galaxy contribution. To include the while a face-on observer sees the scattered light from the upper limits in MAGPHYS we follow the prescription of torus and is not affected by absorption, i.e., a negative da Cunha et al. (2015) in which the flux densities are set τeff (see Siebenmorgen et al. 2015). to zero and the upper limit values are set as the flux uncertainty. 3.4. Fitting Methodology Step 5: We find the total fit by combining the AGN Disentangling the AGN and star-formation contributions to the radio-X-ray emission 11 and the host galaxy component. from the dust clouds along the line of sight. Therefore, a Step 6: We repeat the steps 1–5 above for all the com- comparison of the gray squares with the black plus signs binations of the torus and the accretion disk (95%-65%). indicates the correction of the optical-UV emission from We identify the best-fit as the one that not only results the torus obscuration is negligible. As noted in Section in one of the lowest χ2 values for the total fit but also 3.1.1 the accretion disk templates are built within the one of the lowest χ2 values for individual components. range of ±12◦ from the inclination angle of the best-fit We note that after identifying the best-fit torus model torus model. For this object, the inclination angle from - via χ2 minimization- we examine the impact of varia- the best-fit torus model is 52◦ while from the best-fit tion of the torus parameters on the optical-UV SED and accretion disk template is 63◦. the total fit. This is particularly important for the so- The accretion disk template lies below the X-ray data called “red quasars” in our sample. This class of AGN which suggests the presence of a non-thermal X-ray com- are reddened Type 1 AGN in which the optical-UV SED ponent. Chandra observations Fabian et al. (2003) find lacks the big blue bump and which also have red MIR col- extended X-ray emission on both sides of the nucleus of ors (e.g., Benn et al. 1998; Cutri et al. 2001; Lacy et al. 3C 9 coincident with the radio structure suggesting it is 2004; Georgakakis et al. 2009; Kim & Im 2018). Under- probably due to non-thermal inverse Compton emission standing the nature of red quasars is beyond the scope from the interaction of the relativistic electrons with the of this paper. However, we assume that the reddening is cosmic microwave background. We note that this ex- due to obscuration, and to obtain a reasonable UV bump tended emission does not effect the nuclear X-ray emis- in these sources, we carefully examine the dependence of sion used in our SED modeling. the shape of the intrinsic optical-UV SED on the torus –3C 14 is a red quasars (Smith & Spinrad 1980) with parameters. less than 2% core contribution at 5 GHz and an extended radio jet of 308 kpc. The radio emission is best de- 4. RADIO-TO-X-RAY SED ANALYSIS OF THE 3CRR scribed with a broken power-law with an exponential cut- QUASARS AT 1

Table 6 The parameters of the best-fit torus and accretion disk model

1 2 3 4 5 ◦ 6 ◦ 9 7 8 Name Rin (pc) VC (%) τC τD θT ( ) θAD ( )MBH /M⊙(×10 ) log(λEdd) Spin 3C 009 3.74 1.5 0 1000 52 48±3.5 3.8±1.75 -0.4±0.2 0.98±0.02 3C 014 4.53 77.7 0 100 67 57±3.5 9.6±4.5 -0.4±0.2 0.92±0.02 3C 043 1.80 1.5 45 100 60 70±3.5 2.1±0.96 -1.0±0.2 0.99−0.02 3C 181 2.85 1.5 45 300 52 59±3.5 1.7±0.78 -0.6±0.2 0.92±0.02 3C 186 2.75 1.5 4.5 300 52 62±3.5 1.5±0.69 -0.6±0.2 0.99 −0.02 3C 190 5.95 1.5 45 300 60 63±3.5 6.0±2.8 -1.0±0.2 0.99 −0.02 3C 191 6.53 1.5 45 300 43 43±3.5 1.8±0.82 -0.6±0.2 0.92 ±0.02 3C 204 2.59 77.7 0 300 52 55±3.5 2.9±1.33 -1.0±0.2 0.98 ±0.02 3C 205 5.79 7.7 4.5 300 52 55±3.5 9.8±4.5 -0.8±0.2 0.92 ±0.02 3C 208 2.09 1.5 4.5 300 52 48±3.5 1.5±0.69 -0.4±0.2 0.99 −0.02 3C 212 1.71 7.7 4.5 300 60 53±3.5 6.0±2.8 -1.20±0.2 0.96 ±0.02 3C 245 7.94 1.5 0 30 19 26±3.5 1.1±0.51 -0.6±0.2 0.98 ±0.02 3C 268.4 2.29 1.5 45 1000 52 52±3.5 6.2±2.9 -1.0±0.2 0.99 −0.02 3C 270.1 2.55 1.5 45 1000 52 50±3.5 4.6±2.1 -1.0±0.2 0.98 ±0.02 3C 287 1.16 38.5 0 1000 19 26±3.5 1.5±0.69 -1.2±0.2 0.99 ±0.02 3C 318 1.30 1.5 45 1000 33 38±3.5 2.4±1.1 -1.6±0.2 0.98 ±0.02 3C 325 0.75 1.5 13.5 300 60 49±3.5 3.8±1.75 -1.4±0.2 0.88 ±0.02 4C 16.49 1.19 7.7 0 1000 43 55±3.5 2.1±0.97 -1.4±0.2 0.94 ±0.02 3C 432 3.07 38.5 0 1000 33 46±3.5 4.7±2.2 -1.2±0.2 0.99 −0.02 3C 454.0 4.53 1.5 4.5 1000 43 51±3.5 0.83±0.38 0.00±0.2 0.99 −0.02 11 1) The inner radius of the best-fit torus (Rin = rin × pLAGN /10 ) 2) The clumps volume filling factor 3) The optical depth of the individual clumps 4) The optical depth of the homogeneous disk midplane 5) The inclination angle from the best-fit torus model (in degrees) 6) The inclination angle from the best-fit accretion disk (in degrees) 7) The Eddington ratio, λEdd which is Lbol/LEdd 2 8) The dimensionless spin parameter a ≡ Jc/GMBH where J is the angular momentum of the BH

Table 7 The parameters of the best-fit radio model

ID Best radio model νt νt,jet α1 α2 β νcutoff νcutoff,jet 3C 009 model2 5.5e+07 - -0.09 -1.04 - 5.7e+11 - 3C 014 model2 3.8e+10 - -0.91 -2.15 - 1.0e+13 - 3C 043 model1 - - - -0.75 - 1.5e+12 - 3C 181 model1 - - - -0.94 - 8.1e+11 - 3C 186 model3 5562.3 - - - 10.28 8.3e+12 - 3C 190 model2 4.7e+07 - 1.98 -0.92 - 4.9e+12 - 3C 191 model1 - - - -0.99 - 3.0e+12 - 3C 204 model2 1.1e+07 - 1.37 -1.10 - 1.0e+12 - 3C 205 model2 5.1e+07 - 1.34 -0.98 - 2.1e+11 - 3C 208 model2 1.9e+08 - -0.14 -1.16 - 1.0e+12 - 3C 212 model4 1000.1 1.5e+10 2.20 -0.52 7.46 5.0e+11 7.5e+11 3C 245 model4 3814.4 7.9e+08 2.50 -0.35 9.41 8.1e+12 5.0e+11 3C 268.4 model2 6.9e+09 - -0.71 -1.17 - 3.0e+12 - 3C 270.1 model2 5.6e+07 - 1.35 -0.90 - 1.0e+12 - 3C 287 model2 1.8e+09 - -0.13 -0.79 - 1.3e+12 - 3C 318 model2 4.3e+08 - -0.07 -1.01 - 2.0e+11 - 3C 325 model2 1.7e+09 - -0.50 -1.12 - 2.2e+11 - 4C 16.49 model2 4.2e+08 - -0.40 -1.17 - 1.0e+11 - 3C 432 model2 1.1e+09 - -0.64 -1.19 - 2.8e+11 - 3C 454.0 model3 600.4 - - - 5.92 1.8e+12 - best-fitted accretion disk template also confirms the pres- thermal X-ray component. ence of significant non-thermal X-ray component. –3C 186 is a very well studied CSS (O’Dea 1998) with –3C 181 has a radio jet of 65 kpc with <1% core radio jets of 20 kpc and 4% core contribution at 5 GHz contribution at 5 GHz (Mantovani et al. 1992). The ra- (Spencer et al. 1991; Ludke et al. 1998). The radio emis- dio emission is best described with a single power-law sion is best described with a parabola (a combination of with an exponential cutoff at 814 GHz. Willott et al. multiple power-laws). MERLIN observation at 1.6 MHz (2002) investigate the 850µm band Scuba observation of indicates a one-sided jet and and two bent lobes at 60◦ this source and predict no non-thermal component con- and 90◦ with respect to the source axis, creating a S- tributing to this wavelength. shaped source (Spencer et al. 1991; Ludke et al. 1998). The best-fit torus is a combination of clumps with a The curvature in the lobes is presumably due to the suf- small filling factor and a homogeneous disk with mod- ficiently dense ISM that affect the expansion of the radio erate opacity which results in little correction of the structure. optical-UV emission. The best-fit torus is a combination of clumps with a Comparing the best-fit accretion disk with the X-ray small filling factor and a homogeneous disk with moder- data does not indicate the presence of an underlying non- ate opacity which results in no significant correction of Disentangling the AGN and star-formation contributions to the radio-X-ray emission 13

observed−frame observed−frame 10m 1m 10cm 1cm 1mm 100µm 10µm 1µm 0.1µm 10nm 1keV 10keV 10m 1m 10cm 1cm 1mm 100µm 10µm 1µm 0.1µm 10nm 1keV 10keV 48 48 3C9 3C14 QSO QSO 47 z = 2.01 47 z = 1.47

46 46

45 45 [erg/Hz s]) [erg/Hz s]) ν ν L L ν ν 44 44 log ( log ( Total fit Total fit 43 Best−fit radio model 43 Best−fit radio model Best−fit torus (Siebenmorgen et al. 2015) Best−fit torus (Siebenmorgen et al. 2015) Best−fit accretion disk (Kubota&Done 2018) Best−fit accretion disk (Kubota&Done 2018) Core contribution 0.90% Best−fit host galaxy (MAGPHYS) Core contribution 1.7% Best−fit host galaxy (MAGPHYS)

7 8 9 10 11 12 13 14 15 16 17 18 19 7 8 9 10 11 12 13 14 15 16 17 18 19 log(ν [Hz]) rest−frame log(ν [Hz]) rest−frame observed−frame observed−frame 10m 1m 10cm 1cm 1mm 100µm 10µm 1µm 0.1µm 10nm 1keV 10keV 10m 1m 10cm 1cm 1mm 100µm 10µm 1µm 0.1µm 10nm 1keV 10keV 48 48 3C43 3C181 QSO/CSS QSO 47 z = 1.46 47 z = 1.38

46 46

45 45 [erg/Hz s]) [erg/Hz s]) ν ν L L ν ν 44 44 log ( log ( Total fit Total fit 43 Best−fit radio model 43 Best−fit radio model Best−fit torus (Siebenmorgen et al. 2015) Best−fit torus (Siebenmorgen et al. 2015) Best−fit accretion disk (Kubota&Done 2018) Best−fit accretion disk (Kubota&Done 2018) Core contribution < 5.68% Best−fit host galaxy (MAGPHYS) Core contribution 0.92% Best−fit host galaxy (MAGPHYS)

7 8 9 10 11 12 13 14 15 16 17 18 19 7 8 9 10 11 12 13 14 15 16 17 18 19 log(ν [Hz]) rest−frame log(ν [Hz]) rest−frame observed−frame observed−frame 10m 1m 10cm 1cm 1mm 100µm 10µm 1µm 0.1µm 10nm 1keV 10keV 10m 1m 10cm 1cm 1mm 100µm 10µm 1µm 0.1µm 10nm 1keV 10keV 48 48 3C186 3C190 QSO/CSS QSO/CSS 47 z = 1.07 47 z = 1.20

46 46

45 45 [erg/Hz s]) [erg/Hz s]) ν ν L L ν ν 44 44 log ( log ( Total fit Total fit 43 Best−fit radio model 43 Best−fit radio model Best−fit torus (Siebenmorgen et al. 2015) Best−fit torus (Siebenmorgen et al. 2015) Best−fit accretion disk (Kubota&Done 2018) Best−fit accretion disk (Kubota&Done 2018) Core contribution 4.0% Best−fit host galaxy (MAGPHYS) Core contribution 8.9% Best−fit host galaxy (MAGPHYS)

7 8 9 10 11 12 13 14 15 16 17 18 19 7 8 9 10 11 12 13 14 15 16 17 18 19 log(ν [Hz]) rest−frame log(ν [Hz]) rest−frame observed−frame observed−frame 10m 1m 10cm 1cm 1mm 100µm 10µm 1µm 0.1µm 10nm 1keV 10keV 10m 1m 10cm 1cm 1mm 100µm 10µm 1µm 0.1µm 10nm 1keV 10keV 48 48 3C191 3C204 QSO QSO 47 z = 1.96 47 z = 1.11

46 46

45 45 [erg/Hz s]) [erg/Hz s]) ν ν L L ν ν 44 44 log ( log ( Total fit Total fit 43 Best−fit radio model 43 Best−fit radio model Best−fit torus (Siebenmorgen et al. 2015) Best−fit torus (Siebenmorgen et al. 2015) Best−fit accretion disk (Kubota&Done 2018) Best−fit accretion disk (Kubota&Done 2018) Core contribution 9.3% Best−fit host galaxy (MAGPHYS) Core contribution 8.0% Best−fit host galaxy (MAGPHYS)

7 8 9 10 11 12 13 14 15 16 17 18 19 7 8 9 10 11 12 13 14 15 16 17 18 19 log(ν [Hz]) rest−frame log(ν [Hz]) rest−frame

Figure 3. The full suite of photometry with the best-fit models. The gray data points are the absorption corrected broadband photometry, and the black plus signs in optical-UV bands are the photometric points before the torus obscuration correction. The components are: radio component from the core, jets and lobes (light blue), infrared emission from the torus (dark red), the accretion disk component that accounts for the thermal optical to X-ray emission (green), the host galaxy component obtained from MAGPHYS SED fitting code (magneta) and the total fit (orange). The dark blue star indicates the core contribution to the 5 GHz photometry estimated from the RCD. the optical-UV emission. ies (Hilbert et al. 2016; Siemiginowska et al. 2010). The The Chandra and Hubble Space Telescope (HST ) HST images indicate a blob of star formation activity 2” observations indicate that 3C 186 lives in an over- (corresponding to ∼ 16 kpc at z ∼ 1) away from the cen- dense region which is most likely a cluster of galax- tral engine, perpendicular to the direction of the jets (see 14 Azadi et al.

observed−frame observed−frame 10m 1m 10cm 1cm 1mm 100µm 10µm 1µm 0.1µm 10nm 1keV 10keV 10m 1m 10cm 1cm 1mm 100µm 10µm 1µm 0.1µm 10nm 1keV 10keV 48 48 3C205 3C208 QSO QSO 47 z = 1.53 47 z = 1.11

46 46

45 45 [erg/Hz s]) [erg/Hz s]) ν ν L L ν ν 44 44 log ( log ( Total fit Total fit 43 Best−fit radio model 43 Best−fit radio model Best−fit torus (Siebenmorgen et al. 2015) Best−fit torus (Siebenmorgen et al. 2015) Best−fit accretion disk (Kubota&Done 2018) Best−fit accretion disk (Kubota&Done 2018) Core contribution 3.0% Best−fit host galaxy (MAGPHYS) Core contribution 9.5% Best−fit host galaxy (MAGPHYS)

7 8 9 10 11 12 13 14 15 16 17 18 19 7 8 9 10 11 12 13 14 15 16 17 18 19 log(ν [Hz]) rest−frame log(ν [Hz]) rest−frame observed−frame observed−frame 10m 1m 10cm 1cm 1mm 100µm 10µm 1µm 0.1µm 10nm 1keV 10keV 10m 1m 10cm 1cm 1mm 100µm 10µm 1µm 0.1µm 10nm 1keV 10keV 48 48 3C212 3C245 QSO QSO 47 z = 1.05 47 z = 1.03

46 46

45 45 [erg/Hz s]) [erg/Hz s]) ν ν L L ν ν 44 44 log ( log ( Total fit Total fit 43 Best−fit radio model 43 Best−fit radio model Best−fit torus (Siebenmorgen et al. 2015) Best−fit torus (Siebenmorgen et al. 2015) Best−fit accretion disk (Kubota&Done 2018) Best−fit accretion disk (Kubota&Done 2018) Core contribution 17.% Best−fit host galaxy (MAGPHYS) Core contribution 66.% Best−fit host galaxy (MAGPHYS)

7 8 9 10 11 12 13 14 15 16 17 18 19 7 8 9 10 11 12 13 14 15 16 17 18 19 log(ν [Hz]) rest−frame log(ν [Hz]) rest−frame observed−frame observed−frame 10m 1m 10cm 1cm 1mm 100µm 10µm 1µm 0.1µm 10nm 1keV 10keV 10m 1m 10cm 1cm 1mm 100µm 10µm 1µm 0.1µm 10nm 1keV 10keV 48 48 3C268.4 3C270.1 QSO QSO 47 z = 1.40 47 z = 1.53

46 46

45 45 [erg/Hz s]) [erg/Hz s]) ν ν L L ν ν 44 44 log ( log ( Total fit Total fit 43 Best−fit radio model 43 Best−fit radio model Best−fit torus (Siebenmorgen et al. 2015) Best−fit torus (Siebenmorgen et al. 2015) Best−fit accretion disk (Kubota&Done 2018) Best−fit accretion disk (Kubota&Done 2018) Core contribution 8.4% Best−fit host galaxy (MAGPHYS) Core contribution 22.% Best−fit host galaxy (MAGPHYS)

7 8 9 10 11 12 13 14 15 16 17 18 19 7 8 9 10 11 12 13 14 15 16 17 18 19 log(ν [Hz]) rest−frame log(ν [Hz]) rest−frame observed−frame observed−frame 10m 1m 10cm 1cm 1mm 100µm 10µm 1µm 0.1µm 10nm 1keV 10keV 10m 1m 10cm 1cm 1mm 100µm 10µm 1µm 0.1µm 10nm 1keV 10keV 48 48 3C287 3C318 QSO/CSS QSO/CSS 47 z = 1.05 47 z = 1.57

46 46

45 45 [erg/Hz s]) [erg/Hz s]) ν ν L L ν ν 44 44 log ( log ( Total fit Total fit 43 Best−fit radio model 43 Best−fit radio model Best−fit torus (Siebenmorgen et al. 2015) Best−fit torus (Siebenmorgen et al. 2015) Best−fit accretion disk (Kubota&Done 2018) Best−fit accretion disk (Kubota&Done 2018) Core contribution < 12.4% Best−fit host galaxy (MAGPHYS) Core contribution < 12.1% Best−fit host galaxy (MAGPHYS)

7 8 9 10 11 12 13 14 15 16 17 18 19 7 8 9 10 11 12 13 14 15 16 17 18 19 log(ν [Hz]) rest−frame log(ν [Hz]) rest−frame

Figure 3 (Cont.). Figure 10 in Hilbert et al. 2016). This blob could be ei- –3C 190 is a red quasar (Ishwara-Chandra et al. 2003) ther from the star formation activity of other members of CSS and has a de-projected radio jet length of 124kpc, the cluster or from the host galaxy itself. In latter case, and less than 1% core contribution at 5GHz. The ra- given that the jets’ direction is perpendicular to the blob, dio images (Spencer et al. 1991; Ludke et al. 1998) indi- the star formation activity is not jet-related. Comparing cate multiple features that require a double-power law the best-fit accretion disk with the X-ray data does not in our fitting procedure (see Table 7). As shown in indicate the presence of an underlying non-thermal X-ray Figure 3, the radio emission cutoff occurs at FIR wave- component. lengths. While this suggests that the non-thermal emis- Disentangling the AGN and star-formation contributions to the radio-X-ray emission 15

observed−frame observed−frame 10m 1m 10cm 1cm 1mm 100µm 10µm 1µm 0.1µm 10nm 1keV 10keV 10m 1m 10cm 1cm 1mm 100µm 10µm 1µm 0.1µm 10nm 1keV 10keV 48 48 3C325 4C16.49 QSO QSO 47 z = 1.13 47 z = 1.30

46 46

45 45 [erg/Hz s]) [erg/Hz s]) ν ν L L ν ν 44 44 log ( log ( Total fit Total fit 43 Best−fit radio model 43 Best−fit radio model Best−fit torus (Siebenmorgen et al. 2015) Best−fit torus (Siebenmorgen et al. 2015) Best−fit accretion disk (Kubota&Done 2018) Best−fit accretion disk (Kubota&Done 2018) Core contribution 0.29% Best−fit host galaxy (MAGPHYS) Core contribution 5.0% Best−fit host galaxy (MAGPHYS)

7 8 9 10 11 12 13 14 15 16 17 18 19 7 8 9 10 11 12 13 14 15 16 17 18 19 log(ν [Hz]) rest−frame log(ν [Hz]) rest−frame observed−frame observed−frame 10m 1m 10cm 1cm 1mm 100µm 10µm 1µm 0.1µm 10nm 1keV 10keV 10m 1m 10cm 1cm 1mm 100µm 10µm 1µm 0.1µm 10nm 1keV 10keV 48 48 3C432 3C454.0 QSO QSO/CSS 47 z = 1.78 47 z = 1.76

46 46

45 45 [erg/Hz s]) [erg/Hz s]) ν ν L L ν ν 44 44 log ( log ( Total fit Total fit 43 Best−fit radio model 43 Best−fit radio model Best−fit torus (Siebenmorgen et al. 2015) Best−fit torus (Siebenmorgen et al. 2015) Best−fit accretion disk (Kubota&Done 2018) Best−fit accretion disk (Kubota&Done 2018) Core contribution 2.5% Best−fit host galaxy (MAGPHYS) Core contribution < 25.3% Best−fit host galaxy (MAGPHYS)

7 8 9 10 11 12 13 14 15 16 17 18 19 7 8 9 10 11 12 13 14 15 16 17 18 19 log(ν [Hz]) rest−frame log(ν [Hz]) rest−frame Figure 3 (Cont.). sion may contribute significantly to the submm wave- cal value of radio quasars with little obscuration (see also lengths (hence SFR estimates based on submm data Brotherton et al. 2001; Willott et al. 2002). should be treated carefully), submm observations are re- 3C 191 spectra indicate the presence of a strong ab- quired for constraining the radio fit. sorbing system (6.1 A)˚ associated with C IV absorption The emission from the torus is best described with line (Anderson et al. 1987) in the vicinity of the radio hot a combination of clumps and a homogeneous disk with spots (Hamann et al. 2001). Studies find that quasars maximum acceptable opacity in the torus library (see with strong associated absorption lines are in the early Table 4). This combination with a small filling factor stage of their development and smaller/younger radio and 43◦ inclination angle result in moderate obscuration structures are more common to have associated absorp- of the optical-UV emission of the accretion disk. While tion lines (e.g., Becker et al. 2000, 2001; Willott et al. quasars spectra, in general, show silicate emission fea- 2002). Presence of these absorbing systems distant from ture, the rest-frame 9–16 µm , spectra of 3C 190 indicate the quasar’s core may be associated with a blowout from the presence of silicate absorption (Leipski et al. 2010). nuclear starburst (Hamann et al. 2001). In our fitting procedure, we examined the torus tem- In most of our sources (see Figure 3) the torus com- plates with silicate absorption, however, none of those ponent dominates the blue side of the FIR bands; in- results in better fit compared to that shown in Figure 3. terestingly in 3C 191, it dominates over the host galaxy Comparing the best-fit accretion disk with the X-ray emission at all the MIR-FIR bands. The Herschel up- data does not indicate the presence of an underlying non- per limits in the FIR allow for the possibility of a weak, thermal X-ray component. FIR, cool dust contribution below the submm waveband –3C 191 was classified as a CSS in some earlier studies where the synchrotron component dominates. We note (Akujor & Garrington 1995; Willott et al. 2002), how- that 3C 191 is classified as a hyperluminous quasar with 13 ever, using the classification of Podigachoski et al. (2015) LFIR > 1.2 × 10 L⊙ in which quasar heated dust is the and radio jet of 81 kpc we do not classify this object as a dominant source of IR radiation (Willott et al. 2002). CSS. 3C 191 has ∼9% core contribution at the 5 GHz and Finally, we note that the best-fitted accretion disk its radio emission is best described as a single power-law model indicates the presence of a non-thermal X-ray with a cutoff at FIR wavelengths (see Table 7). There- component. fore, as illustrated in Figure 3 the submm data and likely –3C 204 has an extended jet of ∼543 kpc and ∼ 8% FIR as well has significant non-thermal contribution. core contribution in the 5 GHz band. The quasar radio The emission from the torus is best described with SED is well fitted with a broken power-law (see Table7). a combination of clumps with low volume filling fac- The VLA 4.9 GHz images show multiple features includ- tor and and a homogeneous disk with moderate opacity ing a bright radio core, hot spots, and a one-sided jet and a inclination angle of 43◦ which implies no signifi- that deflects towards the end (Bridle et al. 1994). There cant obscuration of the accretion disk emission. The low is some evidence of [O II] emission perpendicular to the level of obscuration is consistent with the optical spec- jet axis (Bremer et al. 1992; Bridle et al. 1994) at ∼ 90 tral slope, αopt, of 0.7 (Barthel et al. 1990) that is typi- kpc towards the north and ∼ 45 kpc towards the south. 16 Azadi et al.

The emission from the torus is best described with a with an additional core component (the fourth model in combination of clouds with negligible optical depth and Section 3.1.3). To constrain the radio contribution to the a homogeneous disk with moderate opacity. This combi- FIR we recently obtained the SMA data (see Table 2), nation with a inclination angle of 52◦ results in no sig- which is dominated by the non-thermal radiation from nificant obscuration of accretion disk at optical-UV. the radio structure (Figure 3). The extreme variability Comparing the best-fitted accretion disk model with expected in blazars may not be seen in 3C 245 since our the X-ray data suggests the presence of additional, non- recent SMA observation is consistent with older radio thermal X-ray emission. data (Geldzahler & Witzel 1981). –3C 205 has an extended radio structure of ∼ 251 kpc Consistent with its classification, the best-fitted torus with a core contribution of 3% in the 5GHz band. The indicates a combination of clumps with negligible opac- radio emission is best described as a broken power-law ity and a homogeneous disk with small opacity. This (see Table 7) which is consistent with the presence of combination results in no significant obscuration from the multiple radio structures (Lonsdale & Barthel 1984, the clouds and the homogeneous dusty disk. The 1986). torus inclination angle (19◦, see Table 6) and the ac- The emission from the torus is best described with a cretion disk inclination angle (26◦) are consistent with combination of clumps with small opacity and a homoge- the radio-determined inclination angle (Foley & Barthel neous disk with moderate opacity which result in small 1990; Marin & Antonucci 2016). amount of obscuration of the accretion disk radiation. The best-fitted accretion disk model also indicates the 3C 205 spectra indicate the presence of a strong absorb- presence of significant non-thermal X-ray emission. We ing system (3.21 A)˚ associated with C IV absorption line note that since 3C 245 has nearly a face-on inclination (Anderson et al. 1987). angle and significant underlying non-thermal radiation The best-fitted accretion disk model predicts the pres- at radio/submm and X-ray wavelengths, the non-thermal ence of little non-thermal radiation at X-ray band. continuum may contribute significantly in the IR-optical- –3C 208 has an extended radio jet of ∼ 172 kpc with UV bands as well. Given the significant non-thermal ∼ 10% core contribution at 5 GHz. The radio emission emission and quasar’s dominance at optical-UV bands it is best fitted with a broken power-law that is consistent is unlikely to have a reliable estimate of the host galaxy with the presence of multiple radio features including a properties from our SED fits. bright core and the hot spots (Bridle et al. 1994). 3C –3C 268.4 has a jet of 205 kpc and ∼ 8% core contri- 208 has a one-sided jet that is straight for most of its bution at 5 GHzs. The radio images (Lonsdale & Barthel length but deflects towards the end (Bridle et al. 1994). 1986; Liu et al. 1992) indicate multiple features includ- The emission from the torus is best described with a ing a bright core and double hot spots. The radio emis- combination of clumps with a small optical depth and sion is best fitted with a double power-law model with a homogeneous disk with moderate opacity which result cutoff at FIR wavelengths (see Table 7) resulting in sig- in no significant obscuration at optical-UV wavelengths. nificant non-thermal contribution at submm wavelengths The available HST images (F606W and F140W bands) (Willott et al. 2002). do not indicate any evidence of nearby merging sources The emission from the torus is best described with a (Hilbert et al. 2016). combination of clumps and a homogeneous disk with the The X-ray spectrum fitted by the accretion disk model highest acceptable opacity in the torus library (see Table implies some contribution from a non-thermal compo- 4). However, this combination with a small filling factor nent. and inclination angle of 52◦ result in small obscuration of –3C 212 is a red quasar (Aldcroft et al. 2003) with the accretion disk emission. Similar to 3C 191, 3C 268.4 a radio jet of ∼ 216 kpc and ∼ 17% core contribution is also classified as a hyperluminous quasar with LFIR > 13 at 5 GHz. MERLIN observations at 6 cm and 15 cm 2×10 L⊙ (Willott et al. 2002). Additionally, its spectra (Akujor et al. 1991) indicate multiple features including show strong C IV associated absorption (Anderson et al. a bright core that best fitted with the fourth model in 1987). our fitting procedure (see Section 3.1.3 and Table 7). To 3C 268.4 was reported as a lensed quasar with a constrain the radio fit we used recent ALMA observations foreground cluster at z ∼ 0.35 (Sanitt 1976). Recent (see Table 2) which is dominated by the non-thermal HST observations indicate the presence of a bright star- radiation from the radio structure (Figure 3). forming clump 2.5” from the center and an additional Consistent with its classification as a red quasar our oblong source with both optical and radio emission 0.8” torus model indicates a combination of clumps and a from the center (see Figure 10 in Hilbert et al. 2016). homogeneous disk with moderate optical depth. These Since the IR emission from these structures cannot be re- components result in moderate obscuration of the accre- solved from AGN emission with the current data, the hy- tion disk emission. This target also has X-ray and UV perluminous quasar classification should be treated with absorbers (Aldcroft et al. 2003). caution (see 3C 318 below). The X-ray spectrum fitted by the accretion disk model The X-ray spectrum fitted by the accretion disk model implies the presence of a non-thermal X-ray component. implies no significant contribution from a non-thermal –3C 245 is a moderately beamed quasar (with a jet component. oriented at < 20◦ to the line of sight, Foley & Barthel –3C 270.1 has a jet of ∼374 kpc and a relatively high 1990; Marin & Antonucci 2016) with a ∼ 1067 kpc jet core contribution (22%) at 5 GHz. Multi-frequency ra- and ∼ 66% core contribution at 5 GHz which is signifi- dio images of 3C 270.1 indicate the presence of strong cantly higher than other sources in our sample (see Table core and hot spots (Liu et al. 1992) and a one-sided jet 5). The radio emission is best described as a parabola (Hilbert et al. 2016). The radio emission is best fitted with a double power-law with a cutoff at FIR wave- Disentangling the AGN and star-formation contributions to the radio-X-ray emission 17 lengths resulting in significant non-thermal contribution Table 2 and Barthel & Versteeg 2019) to constrain the at submm wavelengths. radio model. Barthel & Versteeg (2019) used the 2 cm The emission from the torus is best described with a VLA image to subtract the non-thermal emission from combination of clumps with a small filling factor and a the ALMA image at 1mm and estimated that ∼11% of homogeneous disk with the highest acceptable opacity in the total flux at 1 mm has a non-thermal origin. We the torus library. This combination with a 52◦ results in use this estimate to constrain the radio model in our fit- little reddening of the accretion disk emission. ting procedure. However, with this prior estimate, the 3C 270.1 spectra indicate the presence of a strong ab- cutoff in our fitting procedure happens at ∼ 200 GHz sorption complex (> 6.17 A)˚ associated with C IV ab- consequently, we estimate no non-thermal contribution sorption line (Anderson et al. 1987). at 1 mm. Therefore there could be . 10% non-thermal The best-fitted accretion disk model indicates the pres- contribution at submm wavelengths. ence of significant non-thermal emission at X-ray wave- The torus emission is best described with a combina- band. Chandra observations find extended X-ray emis- tion of clumps and a homogeneous disk both with high sion that is co-spatial with the radio lobe and peaks at opacity. However, due to the small filling factor of the the position of the hot spots (Wilkes 2011). The ex- clumps and the 33◦ inclination angle these components tended X-ray emission is not included in the X-ray data do not result in a significant obscurtion of the accretion used for this SED analysis. disk emission. –3C 287 is a CSS with a projected radio jet of ∼ 8 The best-fit accretion disk indicates presence of non- kpc. Because the jet inclination angle for this object is thermal X-ray emission. We note that with only few data unknown, we were unable to estimate its de-projected points at optical, UV, and X-ray range the physical pa- jet length. Also we were unable to find the RCD value rameters driven from the accretion disk model (see Table for this object, therefore adopted the average value of the 6) should be treated with caution. CSS quasars in our sample (Section 3.1.3) which resulted We note that 3C 318 is classified as a hyperlumi- 13 in ∼10% core contribution. The radio emission is mod- nous infrared quasar with LFIR > 10 L⊙ in some eled as a double power-law with a cutoff at submm/FIR earlier studies (Willott et al. 2002, 2007). Recently, wavelengths (see Table 7,) resulting in significant non- Podigachoski et al. (2016b) reported that most of the thermal contribution at submm wavelengths. The VLBI fluxes measured in earlier studies originate in a pair of and MERLIN observations indicates the presence of mul- bright interacting galaxies at z ∼ 0.35. To robustly esti- tiple radio structures including a curving jet (Fanti et al. mate the AGN and host galaxy properties, we used the 1989). fluxes from Table 1 in Podigachoski et al. (2016b). After The emission from the torus is best described with a subtracting the contamination of the nearby source from combination of clumps with a negligible optical depth the photometry, 3C 318 has an SFR of ∼ 320M⊙/yr fac- and a homogeneous disk with the highest acceptable tor of 5 lower than the Willott et al. (2007) estimation. opacity in the torus library (see Table 4). This com- –3C 325 was originally classified as a radio galaxy, bination with the inclination angle of 19◦ (26◦ from the and later was reclassified as a red quasar (Grimes et al. accretion disk model) result in no significant obscura- 2005) based on optical spectroscopic data. This quasar tion of the optical-UV radiation from the accretion disk. has a projected radio jet of 124 kpc and less than 1% core While the extreme variability is expected in moderately contribution at 5 GHz. Because the jet inclination angle beamed quasars, it may not be seen in CSS and GPS for this object is unknown, we were unable to estimate its sources such as 3C 287 since they are young (e.g., Conway de-projected jet length. The VLA images show multiple 2002; Salvesen et al. 2009). features including bright hot spots and asymmetrically The X-ray spectrum fitted by the accretion disk model placed lobes (Fernini et al. 1997). The radio data is best implies a significant contribution from a non-thermal fitted with a double power-law which turns down before component. The X-ray data obtained with XMM- reaching the submm wavelengths. Newton and Chandra (Salvesen et al. 2009; Wilkes et al. Consistent with its classification the emission from the 2013) find a soft X-ray spectrum (Γ=1.8) that can be torus is best described with a combination of clumps and fitted with a simple power-law. a homogenous disk with moderate opacity. This com- Similar to 3C 245, 3C 278 is viewed almost face-on bination results in moderate obscuration of the accre- and has significant underlying non-thermal radiation at tion disk emission. Considering the lack of data in 1– radio/submm and X-ray wavelengths. The non-thermal 3 µm (observed-frame) the best-fit accretion disk is de- continuum may contribute at other wavelengths. We also termined with the data points at shorter wavelengths, note that the HST images indicate a few nearby sources therefore the accretion disk parameters are not well con- (∼ 5”, Hilbert et al. 2016) and the IR emission may be strained. contaminated by one/more of these sources. All together, The best-fit accretion disk does not imply presence of these make the host galaxy properties estimated from our non-thermal X-ray emission. The X-ray analysis indi- SED fits uncertain. cates a moderately hard X-ray spectrum with the hard- –3C 318 is a CSS with a jet of ∼ 8kpc and core contri- ness ratio, (H-S/H+S, where H and S are the net count bution < 12% at 5 GHz. The 18 cm MERLIN and VLBI rates in the 2–8 keV and 0.5–2 keV X-ray bands) of +0.05 observations (Spencer et al. 1991) show a two-sided jet which is harder than all the other quasars in our sample which fades before reaching the lobes. The emission from (Wilkes et al. 2013). the radio structure is best described with a double power- –4C 16.49 has a radio jet of 216 kpc with a 5% core law with a cutoff at 200 GHz. In addition to the 1.2 mm contribution at 5 GHz. The radio images obtained with MOMBO data we used recent ALMA observations (see VLA at 2 and 6 cm show a strong radio core, jet and a small counter jet (Lonsdale et al. 1993). The radio emis- 18 Azadi et al. sion is best described with a double power-law (see Table tamination (> 70%) at 1 mm. To robustly estimate the 7). Due to the small number of reliable data points and SFR based on submm observations it is important to lack of submm data we should be cautious in interpreting subtract any source of contamination from the broad- the optimization results and interpretation of the non- band photometry. thermal contribution from the radio structure at shorter – In 13 sources (3C 9, 3C 43, 3C 191, 3C 204, 3C wavelengths. 208, 3C 212, 3C 245, 3C 268.4, 3C 270.1, 3C 287, 3C The best-fit torus model is a combination of clumps 318, 4C 16.49 and 3C 454.0) the best-fit accretion disk with negligible opacity and a homogeneous disk with the confirms the presence of a significant non-thermal X-ray highest acceptable opacity in the torus library (see Table component. 4). This combination results in no significant obscuration – In sources with small inclination angles (3C 245 and of optical-UV emission from the accretion disk. 3C 287) the significant underlying non-thermal radiation The best-fitted accretion disk template also confirms at radio/submm and X-ray wavelengths, may contribute the presence of significant non-thermal X-ray component. at IR-optical-UV bands as well. Therefore, the torus and However, unlike other sources in our sample, 4C 16.49 do the host galaxy properties derived from SED fitting are not have SDSS or other recent reliable data at optical-UV likely to be uncertain. wavelengths and is only limited to old SuperCOSMOS – The HST (F606W and F140W bands, Hilbert et al. observations. Given the limited number of optical-UV 2016) of some of our sources (3C 268.4, 3C 287 and may photometric measurements, the accretion disk may not be 3C 432) indicate the presence of a few nearby objects be well-constrained and the modeled non-thermal X-ray which may contaminate the photometry. In these cases component may not be significant. the host galaxy properties estimated from SED fitting –3C 432 has jet of ∼ 155 kpc and core contribution may be overestimated. of ∼ 3% at 5 GHz. The VLA images indicate multiple – Four quasars in our sample, 3C 191, 3C 205, 3C structures including radio lobes, bright hot spots, and 268.4, 3C 270.1, with de-projected radio jets of 81, 251, a one-sided jet (Bridle et al. 1994). The emission from 205 and 374 kpc have associated C IV absorption com- these radio structures is best described with a double plexes (Anderson et al. 1987) with equivalent width of power-law (Table 7). 6.12 A˚ , 3.21 A,˚ > 1.87 A˚ and > 6.17 A,˚ respec- The torus emission is best fitted with a combination of tively. While some studies (e.g., Becker et al. 2000, 2001; clumps with negligible opacity, and a homogeneous disk Willott et al. 2002) suggest that associated absorption ◦ with high opacity that are viewed at 33 and result in is more common in sources with smaller/younger radio no significant obscuration of the accretion disk emission. structures, there does not appear to be such a relation- The HST F606W and F140W images show an extended ship in our sample. narrow line regions (within 8”) along the direction of the – Our SED modeling is limited by the number of re- radio lobes (Hilbert et al. 2016) with several faint sources liable photometry (see Section 5.1). In sources with few within this radius which could potentially contaminate reliable optical-UV data points (3C 14, 3C 318, 3C 325 the quasar SED. and 4C 16.49) the accretion disk parameters may not be The X-ray spectrum fitted by the accretion disk implies well constrained. no contribution from a non-thermal component to the X- 5. ray emission. RESULTS AND DISCUSSION –3C 454.0 is a CSS with a de-projected jet of < 46 In this study we present a state-of-the-art AGN radio- kpc and < 25% core contribution at 5 GHz. The radio to-Xray SED model (ARXSED) that decomposes radio- images show various features, including the core and hot loud quasars into their AGN and host galaxy components spots (Spencer et al. 1991; Ludke et al. 1998). The ra- by fitting their photometry over ten decades in frequency dio emission is best described with a parabola having a space. With ARXSED we investigate the properties of a cutoff at FIR wavelengths. To constrain the radio model sample of 20 3CRR quasars at 1

5

0 1×109 5×109 1×1010 -2.0 -1.5 -1.0 -0.5 0.0 0.5 0.80 0.85 0.90 0.95 1.00 0 20 40 60 80 λ θ MBH/MO • log ( Edd ) spin (degree)

15 Number of quasars Median: 2.84.5 Median: 1.57.7 Median: 4.545.0 Median: 3001000 1.8 1.5 0.0 300 10

5

0 0 2 4 6 8 10 0 20 40 60 80 100 0 10 20 30 40 50 0 200 400 600 800 1000 τ τ Rin (pc) VC (%) C D Figure 4. The distribution of the physical properties of the AGN derived from the accretion disk and the torus fits for our 3CRR quasars, along with the median values in each case and their associated 25th-75th percentile ranges. lations such as MBH − σ (e.g., Hunt 2003) , Lhost − data and high S/N Herschel observations may result LAGN (Vanden Berk et al. 2006) or color-color diagnos- in underestimation of the non-thermal contamination in tics (e.g., Grewing et al. 1968; Sandage 1971; Elvis et al. some cases (e.g., 3C204), and a corresponding overesti- 2012) to account for the host contribution . Considering mate of SFR. the uncertainty and large scatter in each of these scal- ARXSED is based on the assumption that the pho- ing relations, the physical properties of the AGN derived tometry is dominated by the radiation from the AGN at from their analysis may not be robust. most wavelengths (except for the FIR; see Section 3.4), Like most modern SED fitting codes, ARXSED im- therefore we may underestimate the stellar mass in some plements a self-consistent approach to dust attenuation, sources. Also in sources with small inclination angles in which the intrinsic SED of radio-loud quasars is ap- (3C245 and 3C287, see Table 6 ) beamed non-thermal propriately corrected for the reddening and absorption emission at IR-optical-UV wavelengths, unaccounted in occurring in the torus, the host galaxy, and along the our modeling, add uncertainty to the derived host galaxy line of sight in the Milky Way. properties. We will discuss the results and the limita- The main limitations to our approach are from a lack tions of our technique for SFR and stellar mass measure- of reliable photometry. Several sources in our sample ments in detail in a following paper (Azadi et al. 2020b, (e.g., 3C 325, 4C 16.49) lack reliable optical-UV photom- in prep). etry and/or have very few data points, leaving their ac- cretion disk parameters poorly constrained. 5.2. Physical Properties of the SMBHs ARXSED uses the thermal optical-to-X-ray contin- and Dusty Tori in 3CRR Quasars uum fitting (Zhang et al. 1997) technique to constrain The distributions of the SMBH, torus and accretion the SMBH and the accretion disk properties. Constrain- disk properties derived from the SED fits for our 3CRR ing the accretion disk parameters without observations quasars are shown in Figure. 4, along with the median around the peak of the thermal continuum is challeng- values in each case. The MBH , log(λEdd) and spin are ing. The peak of the accretion disk is sensitive to the BH constrained by the best fit accretion disk (see Table 6). mass, Eddington ratio and the spin (see Figure 1), there- The inner radius of the torus, the volume filling factor, fore not having good constraints results in larger uncer- optical depths of the dust clouds, and the homogeneous tainties. To accurately determine the peak of this ther- dusty disk are constrained by the torus model. The in- mal continuum (occurring around 100-1000A˚ rest-frame clination angle (measured from the pole) is the average for SMBHs) we require far-UV (FUV) observations from value driven from the best fit torus and the accretion space (e.g., the Cosmic Origins Spectrograph on HST ), disk. and soft X-ray data contributing to the high energy part The average BH mass of the quasars in our sample of the thermal continuum. is 3.7e+9 ± 2.7e+9 (standard deviation) M⊙. While Another advantage of ARXSED is that it implements the BH masses in our analysis are obtained from the multi-component radio models that account for a steep- best-fit accretion disk templates, the templates are built ening or cut-off due to the aging of the electron popu- based on initial mass estimated from either C IV λ1548 α lations. We find that a single power-law (Lν ∝ ν ) can or Mg II λ2800 emission lines (Section 3.1.1). We used not adequately model the radio emission when compact this mass as an initial estimate for building the QSOSED structures like cores and hot spots are present. In addi- templates and the log(MBH ) is allowed to vary within tion, the long-wavelength radio photometry usefully con- ± 1 dex of this initial estimate (see Table 3). Using strains the non-thermal radiation from the radio struc- C IV λ1548, Mg II λ2800 or Hβ λ4863, McLure et al. tures at shorter wavelengths. However, a lack of submm (2006) estimated MBH for 18 out of 20 quasars in our 20 Azadi et al. 70 cc: −0.44 cc: 0.21 2.0 pvalue: 0.05 pvalue: 0.4 56 1.5 42 1.0 at 5 GHz

(degrees) 28 CD jet θ R 0.5 14 0.0 0 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 θ θ torus,AD (degrees) torus,AD (degrees)

Figure 5. Left: The relation between the radio core dominance RCD and the average inclination angle of the torus and accretion disk. Quasars with smaller torus/accretion disk inclination angles have higher core fractions. Right: The relation between the jet inclination angle and the average inclination angle of the torus and accretion disk. There is no significant correlation between the two angles. sample (except for 3C 318 and 3C 325) and obtained an We note that although the continuum fitting method average of 3.8e+9 ± 2.9e+9 M⊙. Indeed our estimates can reliably estimate the spin of stellar-mass BHs, its of the 3CRR BH masses are consistent with the BH mass application to AGN can be more challenging. In AGN, estimates of the non-3C radio-loud AGN at similar red- the inner accretion disks are relatively cool (T ∼ 105 K) shifts (e.g. Liu et al. 2006). Recently, Collinson et al. with the bulk of the thermal emission occurring in the (2015, 2017) investigated the intrinsic NIR to X-ray SED far-or-extreme UV regime which is hard to observe from of 11 radio-quiet quasars at 1.6 < z < 2.2. Using the ground. However, recent studies of SMBHs with well- Hα λ6565 line they estimate they estimate an average constrained masses find that spins estimated with the BH mass of 1.5e+9 ± 1.4e+9 M⊙. With a larger sample thermal continuum fitting method are in good agreement at z < 0.5, McLure & Dunlop (2002) find that radio- with estimates from reliable techniques such as the X-ray loud quasars have larger BH masses than their radio- reflection technique (e.g., Capellupo et al. 2016). quiet counterparts although with a large overlap. Laor The best-fit torus parameters indicate low obscuration, (2000) also find a distinction between the BH mass of consistent with nature of the sources (quasars) in our the radio-loud and radio-quiet PG quasars with former sample. Indeed all the sources with high filling factor in having BH mass >1e+9 M⊙. Table 6 have clumps with negligible optical depth. As The average quasar log(λEdd) is −0.87± 0.41 using expected for red quasars (3C14, 3C190, 3C212, 3C325) the QSOSED templates of Kubota & Done (2018). How- the obscuration of the optical-UV emission from the ac- ever, lack of the FUV and soft X-ray data can bias the cretion disk is more pronounced (Figure 3). Also, the measurements towards more massive black holes and/or negligible optical depth of the clumps in 7 out of 20 lower Eddington ratios (Section 5.1). Adopting the quasars in our sample (Table 6) indicates that the pri- SMBHs masses from McLure et al. (2006), and the bolo- mary source of torus obscuration is dust in the toroidal metric correction from Heckman et al. (2004) for esti- disk rather than the clumps. The average inclination an- mating Lbol from L[OIII], recently Daly (2019) estimated gle (measured from the pole) obtained from the best fit ◦ ◦ the average log λEdd for 15 of the quasars in our sam- torus and accretion disk is 49 ± 12 , which is consis- ple to be −0.32± 0.42. Given the scatter in the aver- tent with other studies of non-blazar radio-loud (broad- age values and different methodology used in estimating lined) AGN (e.g., Willott et al. 2000; Arshakian 2005; log(λEdd) the differences between their findings and ours Marin & Antonucci 2016). Also Figure 4 indicates that are not significant. Adopting the accretion disk model of we are not fitting sources with inclination angles < 20◦, Done et al. (2012), OPTXAGNF, Collinson et al. (2015) as these would be blazars. ± estimate an average of log(λEdd) of 0.02 0.57 which is 5.3. not significantly different than the average for our sam- The Orientation of 3CRR Quasars ple. Indeed, Figure 1 in Kubota & Done (2018) illus- We investigate the relation of the radio core domi- trates the fairly small differences in shape and normal- nance, RCD (Section 3.1.3), at 5 GHz and the jets’ in- ization between the QSOSED and OPTXAGNF models. clination angle with the average inclination angle of the Another parameter estimated from our accretion disk torus and accretion disk obtained from our best fit in modeling (i.e., thermal continuum fitting) is the SMBH Figure 5. In each panel we report the correlation coeffi- spin. The average spin of the quasars estimated from cient and its significance level based on r-correlate 11 the best-fit accretion disk is 0.97 ± 0.04. Adopting the routine in IDL. OPTXAGNF accretion disk model (Done et al. 2012) In the left panel of Figure. 5 a statistically significant Collinson et al. (2017) find their sample quasars have correlation is found where quasars with more face-on spins < 0.9. Adopting Blandford & Znajek (1977) (smaller) inclination angles have higher core fractions. framework in which the BH spin is related to the mag- 11 netic field, Daly (2019) estimates the spin of 750 SMBHs r-correlate computes the Spearman’s rank correlation coef- ficient (cc) and the significance of its deviation from zero (pvalue). and for the 15 quasars included in our sample find an A correlation is considered significant if pvalue < 0.05. In this case average spin of 0.99±0.01 (see also Daly 2011). it is unlikely for the correlation to have occurred by accident. Disentangling the AGN and star-formation contributions to the radio-X-ray emission 21

However, 3C 245, the quasar with the highest RCD and limits (especially at 70, 160 and 250 µm) as detections core contribution (66%) in our sample plays an impor- and report the SFR estimated from these bands as up- tant role in deriving this correlation. The trend we find per limits while we do treat them as upper limits. Fi- is as expected based on the Unification model in which nally, unlike ARXSED, Podigachoski et al. (2015) do not the flux of the beamed, jet component decreases as the consider non-thermal radiation from the radio structures inclination angle increases. We also examined the rela- in the submm/FIR bands. Overall, these differences re- tion between RCD and all the other torus parameters sult in lower SFR estimates from our fitting procedure (e.g., filling factor, optical depth of the clouds or disk) than from Podigachoski et al. (2015). We discuss the and the accretion disk parameters and did not find any host galaxy properties (e.g., SFR, stellar mass) and the significant correlation. Given the limited range of RCD discrepancies with Podigachoski et al. (2015) more fully of the quasars in our sample (∼ 0 − 0.4 without 3C 245), in a following paper (Azadi et al. 2020b, in prep). in order to see the variation of the radio core dominance Podigachoski et al. (2016a) improved upon the ap- and obscuration with the inclination angle we need to proach of Podigachoski et al. (2015) by adopting include the edge-on sources. This will be addressed in the Siebenmorgen et al. (2015) torus model and the a later paper, in which we will also compare our SED PEGASE´ SED model (Fioc & Rocca-Volmerange 2019) fits with the clear orientation-dependence of the X-ray for the host galaxies. However, their improved SED obscuration in this sample (Wilkes et al. 2013). model, which like ARXSED treats the submm-to- The right panel of Figure 5 shows the relationship UV SED of the torus, was only run for 12 of the between the radio jet inclination angle (taken from Podigachoski et al. (2015) sample sources. Marin & Antonucci 2016, see Table 5) and the average torus/accretion disk inclination angle. Our results in- 5.5. dicate that the orientation of these structures do not The Median SED of 3CRR Quasars always line up. We speculate that changes in accre- Figure 6 presents the median AGN SED (left) and tion disk/torus orientation coming after jet launching the medians for the individual components (right) in our may ruin any alignments between the two. Therefore study. The shaded region around each component shows the difference is expected to be more noticeable with 25th-75th percentile ranges. Since the number of data longer/older jets. A more definitive study will be car- points between the radio and FIR is very limited in our ried out when we include the radio galaxies which have sources the scatter around the median is is significantly with a wider range of orientation. larger in these bands. The gray curve (in the left panel) 5.4. indicates the median SED of the radio-loud quasars in Comparisons with the Literature on the SED Elvis et al. (1994) normalized at 1.5 µm and the dotted Fitting of 3CR Sample at 1

observed−frame observed−frame 10m 1m 10cm 1cm 1mm 100µm 10µm 1µm 0.1µm 10nm 1keV 10keV 10m 1m 10cm 1cm 1mm 100µm 10µm 1µm 0.1µm 10nm 1keV 10keV 48 48 Median AGN fit Median radio fit 47 47 Median torus fit Median accretion disk fit Median host galaxy fit 46 46 [erg/s]) [erg/s])

ν 45 ν 45 L L ν ν 44 44 log ( log ( 43 43 Elvis et al. 1994 (RL) Podigachoski et al. 2015

7 8 9 10 11 12 13 14 15 16 17 18 19 7 8 9 10 11 12 13 14 15 16 17 18 19 log(ν [Hz]) rest−frame log(ν [Hz]) rest−frame Figure 6. Left: The median SED of the total AGN fit in our sample. The gray curve shows the median SED of the radio-loud quasars in Elvis et al. (1994) normalized at 1.5 µm. The dotted gray curve indicates regions with few or no data available in Elvis et al. (1994) sample. Right: The median SED of the each of the components used in this study. The dotted-dashed curve is the median SED of Podigachoski et al. (2015) which is obtained with the AGN and host galaxy components combined. The shaded region around each component indicates the 25th-75th percentile ranges. The small number of photometric measurements between the radio and FIR bands results in a larger scatter around the median in these bands. dian SED cannot plausibly be said to capture the average component from UV to radio wavelengths (MAGPHYS behavior of radio-selected AGN at z > 1. da Cunha et al. 2008, 2015). Figure 6 also shows the median SED of the quasars We have compiled the radio to X-ray SED of 20 (and their host galaxies) from Podigachoski et al. (2015). radio-loud quasars from the 3CRR sample at 1 < Given the scatter in the Podigachoski et al. SEDs (0.5-1 z . 2 by combining archival multi-frequency radio ob- dex, not shown for clarity), their median SED is gener- servations, recent SMA/ALMA observations, Herschel, ally consistent with ours. One notable difference is the WISE, Spitzer, 2MASS, UKIRT, SDSS, XMM-Newton redder and fainter FIR peak relative to our sample. This and Chandra for our analysis (Section 2). In order to could be due to the difference in the torus models i.e., obtain the intrinsic SED of the AGN in our sample, we the larger grains adopted in ARXSED, which results in correct the photometry for the reddening and absorption stronger FIR radiation from the torus and consequently in the host galaxy, the Milky Way as well as the dusty less pronounced emission from the host galaxy. We also torus (Section 3.3). note that ARXSED subtracts the non-thermal radiation In this paper we present the fitting results for individ- from radio structures in the submm/FIR regime, which ual sources (Section 4.1) and analyze the physical prop- can contribute to a fainter FIR peak. erties of the AGN components derived from our modeling There have been many attempts to describe the av- (Section 5.2). Our main findings are as follows: erage SED of quasars (e.g. Richards et al. 2006). Con- • α sidering that the Richards et al. (2006) median SED is A simple power-law (Lν ∝ ν ), is unable to repli- dominated by radio-quiet AGN (their sample includes cate the radio emission from our sources when com- only 8 radio-loud sources) it is not shown in Figure 6. plex radio structures (i.e., lobes, jets, cores, hot spots) are present (Section 4.1). 6. SUMMARY • We predict that in half of our quasar sample there In this study we present a state-of-the-art AGN radio- is a significant (> 70%) non-thermal contribution to-X-ray SED model (ARXSED) that simultaneously fits at submm wavelengths. To robustly estimate the AGN and the host galaxy components. Using this model SFR, when including submm observations in the we fit the radio-to-X-ray SED of a sample of 20 radio- fit, it is important to subtract this source of con- loud quasars from the 3CRR sample at 1

with those in the literature for similarly defined the Max Planck Society, and the Higher Education samples (Section 5.4). Funding Council for England. The SDSS Web Site is http://www.sdss.org/. The SDSS is managed by the • The average inclination angles from the best fit ◦ ◦ ◦ ◦ Astrophysical Research Consortium for the Participat- torus and accretion disk are 48 ±13 and 50 ±11 , ing Institutions. The Participating Institutions are the respectively. The average inclination angle of the American Museum of Natural History, Astrophysical In- radio jets reported in the literature for our sam- ◦ ◦ stitute Potsdam, University of Basel, University of Cam- ple is 33 ± 14 . We do not find a statistically bridge, Case Western Reserve University, University of significant correlation between the jet inclination Chicago, Drexel University, Fermilab, the Institute for angle and the inclination angle from our best fit Advanced Study, the Japan Participation Group, Johns torus/accretion disk model. We speculate that Hopkins University, the Joint Institute for Nuclear As- this misalignment could occur after the jets are trophysics, the Kavli Institute for Particle Astrophysics launched (Section 5.3). and Cosmology, the Korean Scientist Group, the Chi- • The quasars in our sample have a limited range nese Academy of Sciences (LAMOST), Los Alamos Na- of the radio core dominance, R . To investigate tional Laboratory, the Max-Planck-Institute for Astron- CD omy (MPIA), the Max-Planck-Institute for Astrophysics the relation of the RCD with the torus or accretion disk parameters, edge-on sources with wider range (MPA), New Mexico State University, Ohio State Univer- sity, University of Pittsburgh, University of Portsmouth, of RCD should be considered as well. We defer any conclusion until we have fit the SEDs of the ra- Princeton University, the United States Naval Observa- dio galaxies and their best fit torus/accretion disk tory, and the University of Washington. parameters are available (Section 5.3). This research is based on observations made by Her- schel, which is an ESA space observatory with science • We present the median intrinsic SED of the radio- instruments provided by European-led Principal Inves- loud quasars at 1 < z . 2 . We find that the tigator consortia and with important participation from median SED of Elvis et al. (1994), obtained based NASA. This work is based in part on observations made on a sample of radio-loud quasars at z < 1 can with the Spitzer Space Telescope, which was operated not describe the SED of the radio-selected AGN at by the Jet Propulsion Laboratory, California Institute of z > 1. The difference between the two median SED Technology under a contract with NASA. could be due to the sample selection, redshift and We acknowledge the use of Ned Wright’s calculator observations limitaion (Section 5.5). (Wright 2006) and NASA/IPAC Extragalactic Database (NED), operated by the Jet Propulsion Laboratory, Cal- • Our SED models successfully reproduce the ob- ifornia Institute of Technology, under contract with the served photometry and constrain the parameters National Aeronautics and Space Administration. describing the structures surrounding SMBHs at We acknowledge the use of IRAF which is distributed z > 1. Given the large number of parameters, di- by the National Optical Astronomy Observatories, and verse data quality, and possible variability which operated by the Association of Universities for Research may bias our SED fits, we apply priors to the fits in Astronomy, Inc., under contract to the National Sci- based on independent measurements of parameters ence Foundation. such as MBH from the literature. This helps to en- sure consistency of the SED fitting results which REFERENCES otherwise may not be unique. Aird, J., et al. 2012, ApJ, 746, 90 ACKNOWLEDGEMENTS Akujor, C. E., & Garrington, S. T. 1995, A&AS, 112, 235 Akujor, C. E., Spencer, R. E., & Saikia, D. J. 1991, A&A, 249, 337 Support for this work was provided by the National Akujor, C. E., et al. 1994, A&AS, 105, 247 Aeronautics and Space Administration and the Chandra Aldcroft, T. L., et al. 2003, ApJ, 597, 751 X-ray Center (CXC), which is operated by the Smithso- Anderson, S. F., et al. 1987, AJ, 94, 278 Arshakian, T. G. 2005, A&A, 436, 817 nian Astrophysical Observatory for and on behalf of the Aslanian, A. M., et al. 1968, Astrofizika, 4, 129 National Aeronautics Space Administration under con- Azadi, M., et al. 2015, ApJ, 806, 187 tract NAS8-03060 (BJW, MAz,JK). Azadi, M., et al. 2017, ApJ, 835, 27 The scientific results in this article are based to a signif- Barthel, P., & Versteeg, J. 2019, The Messenger, 176, 37 Barthel, P. D., Tytler, D. R., & Thomson, B. 1990, A&AS, 82, icant degree on observations made by the Chandra X-ray 339 Observatory (CXO). This research has made use of data Baskin, A., & Laor, A. 2005, MNRAS, 356, 1029 obtained from the Chandra Data Archive. Becker, R. H., et al. 2000, ApJ, 538, 72 This research has made use of data provided by the Na- Becker, R. H., et al. 2001, ApJS, 135, 227 tional Radio Astronomy Observatory which is a facility Benn, C. R., et al. 1998, MNRAS, 295, 451 Bernhard, E., et al. 2016, MNRAS, 460, 902 of the National Science Foundation operated under coop- Berta, S., et al. 2013, A&A, 551, A100 erative agreement by Associated Universities, Inc. and Best, P. N., & Heckman, T. M. 2012, MNRAS, 421, 1569 data from the Sloan Digital Sky Survey (SDSS). Fund- Best, P. N., et al. 2005, MNRAS, 362, 25 ing for the SDSS and SDSS-II has been provided by Best, P. N., et al. 2007, MNRAS, 379, 894 the Alfred P. Sloan Foundation, the Participating In- Bianchi, S., et al. 2009, A&A, 495, 421 Blandford, R. D., & K¨onigl, A. 1979, ApJ, 232, 34 stitutions, the National Science Foundation, the U.S. Blandford, R. D., & Znajek, R. L. 1977, MNRAS, 179, 433 Department of Energy, the National Aeronautics and Boquien, M., et al. 2019, A&A, 622, A103 Space Administration, the Japanese Monbukagakusho, Braatz, J. A., et al. 1993, ApJ, 409, L5 24 Azadi et al.

Braude, S. Y., et al. 1969, MNRAS, 143, 289 Heckman, T. M., & Best, P. N. 2014, ARA&A, 52, 589 Braude, S. Y., et al. 1970, Astrophys. Lett., 5, 129 Heckman, T. M., et al. 2004, ApJ, 613, 109 Bremer, M. N., et al. 1992, MNRAS, 254, 614 Hilbert, B., et al. 2016, ApJS, 225, 12 Bridle, A. H., et al. 1994, AJ, 108, 766 H¨onig, S. F., & Kishimoto, M. 2010, A&A, 523, A27 Brotherton, M. S., et al. 2001, ApJ, 546, 775 H¨onig, S. F., et al. 2013, ApJ, 771, 87 Brown, A., et al. 2019, ApJ, 871, 87 Hunt, A. M. L. K. 2003, ApJL, 589, L21 Bruzual, G., & Charlot, S. 2003, MNRAS, 344, 1000 Ishwara-Chandra, C. H., Dwarakanath, K. S., & Anantharamaiah, Burns, J. O. 1990, AJ, 99, 14 K. R. 2003, Journal of Astrophysics and Astronomy, 24, 37 Cameron, M., et al. 1993, ApJ, 419, 136 Janssen, R. M. J., et al. 2012, A&A, 541, A62 Capellupo, D. M., et al. 2016, MNRAS, 460, 212 Jester, S. 2005, ApJ, 625, 667 Chabrier, G. 2003, PASP, 115, 763 Kellermann, K. I., & Pauliny-Toth, I. I. K. 1973, AJ, 78, 828 Charlot, S., & Fall, S. 2000, ApJ, 539, 718 Kellermann, K. I., Pauliny-Toth, I. I. K., & Williams, P. J. S. Cleary, K., et al. 2007, ApJ, 660, 117 1969, ApJ, 157, 1 Cohen, A. S., et al. 2004, ApJS, 150, 417 Kellermann, K. I., et al. 1989, AJ, 98, 1195 Cohen, A. S., et al. 2007, AJ, 134, 1245 Kellermann, K. I., et al. 2016, ApJ, 831, 168 Colla, G., et al. 1970, A&AS, 1, 281 Kim, D., & Im, M. 2018, A&A, 610, A31 Colla, G., et al. 1972, A&AS, 7, 1 King, A. 2016, MNRAS, 456, L109 Colla, G., et al. 1973, A&AS, 11, 291 Kirkpatrick, A., et al. 2017, ApJ, 849, 111 Collinson, J. S., et al. 2015, MNRAS, 449, 2174 Kishimoto, M., et al. 2011, A&A, 527, A121 Collinson, J. S., et al. 2017, MNRAS, 465, 358 Konigl, A. 1981, ApJ, 243, 700 Condon, J. J., et al. 1983, AJ, 88, 20 Kriek, M., et al. 2009, ApJ, 700, 221 Condon, J. J., et al. 1998, AJ, 115, 1693 Kr¨ugel, E. 2009, A&A, 493, 385 Conway, J. E. 2002, ??jnlNew A Rev., 46, 263 Kr¨ugel, E., & Siebenmorgen, R. 1994, A&A, 288, 929 Coziol, R., et al. 2017, MNRAS, 466, 921 Kubota, A., & Done, C. 2018, MNRAS, 480, 1247 Crummy, J., et al. 2006, MNRAS, 365, 1067 Kuehr, H., et al. 1981, A&AS, 45, 367 Cutri, R. M., et al. 2001, in American Astronomical Society Kuraszkiewicz, J. K., et al. 2002, ApJS, 143, 257 Meeting Abstracts, Vol. 198, American Astronomical Society Kuraszkiewicz, J. K., et al. 2003, ApJ, 590, 128 Meeting Abstracts #198, 33.17 Lacy, M., et al. 2004, ApJS, 154, 166 Czerny, B., et al. 2003, A&A, 412, 317 Laing, R. A., & Peacock, J. A. 1980, MNRAS, 190, 903 da Cunha, E., Charlot, S., & Elbaz, D. 2008, MNRAS, 388, 1595 Laing, R. A., Riley, J. M., & Longair, M. S. 1983, MNRAS, 204, da Cunha, E., et al. 2015, ApJ, 806, 110 151 Daly, R. A. 2011, MNRAS, 414, 1253 Lani, C., Netzer, H., & Lutz, D. 2017, MNRAS, 471, 59 Daly, R. A. 2019, ApJ, 886, 37 Laor, A. 2000, ApJ, 543, L111 Davis, S. W., & Laor, A. 2011, ApJ, 728, 98 Laor, A., & Davis, S. W. 2014, MNRAS, 438, 3024 Done, C., et al. 2012, MNRAS, 420, 1848 Large, M. I., et al. 1981, MNRAS, 194, 693 Douglas, J. N., et al. 1996, AJ, 111, 1945 Lee, S.-K., et al. 2010, ApJ, 725, 1644 Drouart, G., et al. 2014, A&A, 566, A53 Leipski, C., et al. 2010, ApJ, 717, 766 Edelson, R. A., & Malkan, M. A. 1986, ApJ, 308, 59 Leja, J., et al. 2017, ApJ, 837, 170 Efstathiou, A., & Rowan-Robinson, M. 1995, MNRAS, 273, 649 Leja, J., et al. 2018, ApJ, 854, 62 Elvis, M. 2010, in IAU Symposium, Vol. 267, Co-Evolution of Liu, R., Pooley, G., & Riley, J. M. 1992, MNRAS, 257, 545 Central Black Holes and Galaxies, ed. B. M. Peterson, R. S. Liu, Y., Jiang, D. R., & Gu, M. F. 2006, ApJ, 637, 669 Somerville, & T. Storchi-Bergmann, 55–64. [link] Lonsdale, C. J., & Barthel, P. D. 1984, A&A, 135, 45 Elvis, M., et al. 1994, ApJS, 95, 1 Lonsdale, C. J., & Barthel, P. D. 1986, AJ, 92, 12 Elvis, M., et al. 2012, ApJ, 759, 6 Lonsdale, C. J., Barthel, P. D., & Miley, G. K. 1993, ApJS, 87, 63 Fabian, A. C., Celotti, A., & Johnstone, R. M. 2003, MNRAS, Ludke, E., et al. 1998, MNRAS, 299, 467 338, L7 Mack, K. H., et al. 2005, A&A, 435, 863 Fanaroff, B. L., & Riley, J. M. 1974, MNRAS, 167, 31P Madau, P., & Dickinson, M. 2014, ARA&A, 52, 415 Fanti, C., et al. 1974, A&AS, 18, 147 Mantovani, F., et al. 1992, MNRAS, 257, 353 Fanti, C., et al. 1989, A&A, 217, 44 Marin, F., & Antonucci, R. 2016, ApJ, 830, 82 Feltre, A., et al. 2012, MNRAS, 426, 120 Markwardt, C. B. 2009, in Astronomical Society of the Pacific Fernini, I., Burns, J. O., & Perley, R. A. 1997, AJ, 114, 2292 Conference Series, Vol. 411, Astronomical Data Analysis Ficarra, A., Grueff, G., & Tomassetti, G. 1985, A&AS, 59, 255 Software and Systems XVIII, ed. D. A. Bohlender, D. Durand, Fioc, M., & Rocca-Volmerange, B. 2019, A&A, 623, A143 & P. Dowler, 251. [link] Foley, A. R., & Barthel, P. D. 1990, A&A, 228, 17 McLure, R. J., & Dunlop, J. S. 2002, MNRAS, 331, 795 Fritz, J., Franceschini, A., & Hatziminaoglou, E. 2006, MNRAS, McLure, R. J., et al. 2006, MNRAS, 368, 1395 366, 767 Middei, R., et al. 2018, arXiv e-prints, arXiv:1803.07334 Geach, J. E., et al. 2011, ApJ, 730, L19 Mullaney, J. R., et al. 2011, MNRAS, 414, 1082 Geldzahler, B. J., & Witzel, A. 1981, AJ, 86, 1306 Nenkova, M., et al. 2008, ApJ, 685, 147 Georgakakis, A., et al. 2009, MNRAS, 394, 533 Netzer, H., et al. 2007, ApJ, 666, 806 Ghisellini, G. 2013, Radiative Processes in High Energy Neugebauer, G., et al. 1979, ApJ, 230, 79 Astrophysics, Vol. 873. [link] Novikov, I. D., & Thorne, K. S. 1973, in Black Holes (Les Astres Gierli´nski, M., & Done, C. 2004, MNRAS, 349, L7 Occlus), 343–450 Gower, J. F. R., Scott, P. F., & Wills, D. 1967, MmRAS, 71, 49 O’Dea, C. P. 1998, PASP, 110, 493 Gregory, P. C., & Condon, J. J. 1991, ApJS, 75, 1011 Orr, M. J. L., & Browne, I. W. A. 1982, MNRAS, 200, 1067 Grewing, M., Demoulin, M.-H., & Burbidge, G. R. 1968, ApJ, Page, M. J., et al. 2012, Nature, 485, 213 154, 447 Pauliny-Toth, I. I. K., Wade, C. M., & Heeschen, D. S. 1966, Grimes, J. A., Rawlings, S., & Willott, C. J. 2005, MNRAS, 359, ApJS, 13, 65 1345 Petrucci, P. O., et al. 2001, ApJ, 556, 716 Haardt, F., & Maraschi, L. 1993, ApJ, 413, 507 Petrucci, P. O., et al. 2013, A&A, 549, A73 Hales, S. E. G., Baldwin, J. E., & Warner, P. J. 1993, MNRAS, Petrucci, P. O., et al. 2018, A&A, 611, A59 263, 25 Piconcelli, E., et al. 2005, A&A, 432, 15 Hales, S. E. G., et al. 1990, MNRAS, 246, 256 Pier, E. A., & Krolik, J. H. 1992, ApJ, 401, 99 Hales, S. E. G., et al. 1995, MNRAS, 274, 447 Podigachoski, P., et al. 2015, A&A, 575, A80 Hamann, F. W., et al. 2001, ApJ, 550, 142 Podigachoski, P., et al. 2016a, MNRAS, 462, 4183 Hao, H., et al. 2014, MNRAS, 438, 1288 Podigachoski, P., et al. 2016b, A&A, 585, A142 Hardcastle, M. J., & Croston, J. H. 2020, arXiv e-prints, Polletta, M., et al. 2000, A&A, 362, 75 arXiv:2003.06137 Polletta, M., et al. 2007, ApJ, 663, 81 Disentangling the AGN and star-formation contributions to the radio-X-ray emission 25

Rees, M. J. 1984, ARA&A, 22, 471 Spencer, R. E., et al. 1991, MNRAS, 250, 225 Reynolds, C. S. 2019, Nature Astronomy, 3, 41 Spinrad, H., et al. 1985, PASP, 97, 932 Richards, G. T., et al. 2006, ApJS, 166, 470 Steppe, H., et al. 1995, A&AS, 113, 409 Rieke, G. H., & Lebofsky, M. J. 1981, ApJ, 250, 87 Suganuma, M., et al. 2006, ApJ, 639, 46 Roche, P. F., et al. 1991, MNRAS, 248, 606 Sulentic, J. W., et al. 2002, ApJ, 566, L71 Rosario, D. J., et al. 2012, A&A, 545, A45 Svensson, R., & Zdziarski, A. A. 1994, ApJ, 436, 599 Rovilos, E., et al. 2012, A&A, 546, A58 Tody, D. 1986, in Society of Photo-Optical Instrumentation Rowan-Robinson, M. 1995, MNRAS, 272, 737 Engineers (SPIE) Conference Series, Vol. 627, Instrumentation Salvesen, G., et al. 2009, ApJ, 692, 753 in astronomy VI, ed. D. L. Crawford, 733. [link] Sandage, A. R. 1971, in Study Week on Nuclei of Galaxies, ed. Urry, C. M., & Padovani, P. 1995, PASP, 107, 803 D. J. K. O’Connell, 271 Vanden Berk, D. E., et al. 2006, AJ, 131, 84 Sanitt, N. 1976, MNRAS, 174, 91 Viner, M. R., & Erickson, W. C. 1975, AJ, 80, 931 Shakura, N. I., & Sunyaev, R. A. 1976, MNRAS, 175, 613 Waldram, E. M., et al. 1996, MNRAS, 282, 779 Shang, Z., et al. 2011, ApJS, 196, 2 Westhues, C., et al. 2016, AJ, 151, 120 Shangguan, J., et al. 2020, ApJ, 899, 112 Wilkes, B. 2011, The Environmental Impact of the High-redshift Shen, Y. 2013, Bulletin of the Astronomical Society of India, 41, (1.532) Radio-Loud Quasars 3C270.1, Chandra Proposal 61 Wilkes, B. J., et al. 2013, ApJ, 773, 15 Shimmins, A. J., & Wall, J. V. 1973, Willott, C. J., Mart´ınez-Sansigre, A., & Rawlings, S. 2007, AJ, Australian Journal of Physics, 26, 93 133, 564 Siebenmorgen, R., Heymann, F., & Efstathiou, A. 2015, A&A, Willott, C. J., et al. 2000, MNRAS, 316, 449 583, A120 Willott, C. J., et al. 2002, MNRAS, 331, 435 Siebenmorgen, R., Voshchinnikov, N. V., & Bagnulo, S. 2014, Wills, B. J. 1975, Australian Journal of Physics Astrophysical A&A, 561, A82 Supplement, 38, 1 Siebenmorgen, R., et al. 2005, Astronomische Nachrichten, 326, Wright, A., & Otrupcek, R. 1990, PKS Catalog (1990, 0 556 Wright, E. L. 2006, PASP, 118, 1711 Siemiginowska, A., et al. 2010, ApJ, 722, 102 Zhang, S. N., Cui, W., & Chen, W. 1997, ApJ, 482, L155 Slee, O. B. 1995, Australian Journal of Physics, 48, 143 Smith, H. E., & Spinrad, H. 1980, PASP, 92, 553

APPENDIX Table A1, A2 and A3 below presents high, medium and low frequency radio data used in the SED analysis in this study with their references. The fluxes densities are in Jansky. 26 Azadi et al. 0 0.040 0.040 0.040 0.040 0.051 0.040 0.040 0.040 0.040 0.040 0.040 0.040 0.01 0.040 0.061 0.040 0.040 0.040 0.040 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 0.066 0.981 0.075 1.021 0.157 1.698 0.092 1.254 0.095 1.395 0.038 0.586 0.077 0.960 0.033 0.505 0.070 1.122 0.069 0.960 0.102 1.426 0.072 1.072 0.240 2.19 0.111 1.506 0.413 4.651 0.115 1.345 0.099 1.860 0.055 0.768 0.047 0.680 0.107 1.223 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 0 0.345 0.050 0.482 0.030 0.549 0.040 1.173 0.050 0.675 0.060 0.694 0.050 0.308 0.060 0.559 0.030 0.369 0.040 0.688 0.050 0.506 0.040 0.746 0.030 0.608 0.040 1.742 0.040 0.875 0.060 3.106 0.030 0.852 0.040 1.050 0.060 0.407 0.030 0.787 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± L80 G91 L80 L80 G91 L80 L80 G91 L80 L80 G91 L80 L80 G91 L80 L80 G91 L80 L80 G91 L80 L80 G91 L80 L80 G91 L80 L80 G91 L80 L80 G91 L80 L80 G91 W75 L80 G91 L80 L80 G91 L80 L80 G91 L80 L80 G91 L80 L80 G91 L80 L80 G91 L80 W90 G91 W90 0.320 0.546 0.606 1.082 0.655 0.814 0.377 0.457 0.338 0.665 0.536 0.884 0.596 1.380 0.864 0.745 0.824 0.308 0.784 0.18 3.237 ± C83 L80 G91 L80 2.20 z 8085 MHz 5000 MHz 4.85 GHz 2695 MHz Table 1 0.04 ± High frequency radio data 0.019 0.019 0.038 0.019 0.019 0.014 0.019 0.015 0.019 0.019 0.019 0.038 1.04 0.019 0.019 0.047 0.019 0.019 0.015 0.028 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± L80 0.084 0.010 0.216 0.010 0.216 0.010 0.610 0.010 0.291 0.010 0.356 0.010 0.225 0.010 0.159 0.010 0.234 0.010 0.234 0.010 0.497 0.010 0.966 0.010 0.244 0.010 0.450 0.030 1.726 0.010 0.319 0.020 0.394 0.010 0.131 0.010 0.394 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 0.130 0.150 0.200 0.300 0.180 0.110 0.160 0.140 0.480 0.180 0.440 0.230 0.250 0.080 0.300 0.15 0.850 ± G81 L80 L80 S73 0.94 0.042 0.450 0.090 1.420 ± ± 0.010 0.189 0.020 0.512 ± ± SJ95 SJ95 SJ95 SJ95 Name3C 9 230 GHz 90 GHz 31.4 GHz 14900 MHz 10700 MHz 8870 MHz 8400 MH 3C 14 3C 43 0.064 3C 181 3C 190 3C 186 3C 191 3C 204 3C 205 3C 208 3C 212 3C 245 3C 268.4 3C 270.1 3C 287 0.116 3C 318 3C 325 3C 432 4C 16.49 3C 454.0 Disentangling the AGN and star-formation contributions to the radio-X-ray emission 27 1.900 2.700 2.200 2.700 2.700 1.100 3.100 2.400 4.400 2.700 1.600 2.700 4.400 2.200 2.700 1.600 0.643 1.100 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 W96 29.6 21.100 33.100 30.800 22.400 27.400 29.800 41.500 39.200 34.300 26.800 24.200 24.400 35.700 0.699 19.400 0.075 33.200 ± ± H93 L80 0.670 13.98 Hz 151 MHz 86 MHz 3.00 15.53 2.40 20.500 ± ± ± L80 0.632 1.199 0.850 0.872 20.3 0.741 0.561 0.785 0.823 1.008 1.373 0.572 0.709 0.768 15.59 0.818 1.264 16.3 0.790 1.134 0.84 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± L80 L80 L80 L80 K81 W96 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 K81 L80 L80 L80 G67 0.085 13.407 ± D96 Table 2 9.212 0.295 17.004 0.282 17.440 0.178 20.165 0.054 12.644 0.116 11.990 0.098 10.5 0.134 14.824 0.119 11.227 0.210 15.696 0.151 16.459 0.200 13.734 0.166 11.445 0.083 14.170 0.109 15.369 0.090 16.350 0.084 12.644 0.156 15.805 0.086 11.336 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± Medium frequency radio data D96 D96 9.276 12.124 0.955 14.959 0.41 9.742 0.18 6.485 0.15 6.484 0.17 6.268 0.12 6.585 0.37 9.449 0.22 8.337 0.25 10.332 0.32 5.491 0.32 7.485 0.11 6.582 0.24 9.088 0.515 8.556 0.21 7.624 0.18 6.830 0.32 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± L81 D96 L81 D96 L81 D96 L81 D96 L81 D96 L81 D96 L81 D96 L81 D96 7.485 L80 L81 D96 L81 D96 L81 D96 L81 F85 D96 F74 D96 F85 D96 C72 D96 C70 D96 5.36 0.183 ± 0.170 5.65 0.167 6.05 0.194 0.0807 4.440 0.251 11.942 0.09 8.200 0.174 5.69 0.178 8.90 0.183 7.01 0.175 7.75 0.176 0.060 7.32 0.170 7.32 0.167 5.55 0.174 7.63 0.182 6.440 0.171 6.73 0.206 5.85 0.189 7.79 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 0.691 3.760 0.072 3.460 0.0511 0.0557 2.900 0.1253 6.190 0.072 2.6885 0.2116 9.670 0.072 5.04 0.0594 3.620 0.0992 5.060 0.0711 4.440 0.0709 4.560 0.072 3.950 0.082 2.370 0.0555 3.440 0.0371 2.740 0.072 4.300 0.082 4.280 0.103 3.580 0.123 3.890 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± L80 L80 L80 C98 L80 L80 P66 L80 L80 L80 P66 L80 L80 L80 L80 L80 L80 L80 L80 C98 L80 C98 L80 C98 C90 L80 C98 L80 C98 L80 C98 L80 C98 L80 C98 L80 C98 L80 C98 L80 3C 454.0 2.140 4C 16.49 1.4565 3C 432 1.5792 3C 325 3.5645 3C 318 2.562 3C 287 7.0526 3C 270.1 2.727 3C 268.4 1.9804 3C 245 3.3062 3C 212 2.3713 3C 208 2.3648 3C 205 2.387 3C 204 1.297 3C 191 1.8498 3C 181 2.3044 3C 186 1.2369 3C 190 2.552 3C 43 2.819 3C 14 1.873 Name3C 9 1400 MHz 1.955 750 MHz 635 MHz 408 MHz 365 MHz 326MHz 178 MHz 160 M 28 Azadi et al. 39.600 36.0 66.000 42.000 42.000 ± ± ± ± ± L80 L80 192.000 204.000 84 84 276.000 47 160 45.1 170.400 32.3 162.0 ± ± ± ± ± ± ry & Condon (1991), H90 - cock (1980), L81 - Large et al. 975), W90 - Wright & Otrupcek t al. (1983), C98 - Condon et al. 66 380. 61 45 120 120 320. 27.7 161 28.6 147 ± ± ± ± ± ± 32 300. 35 132 65 210 35 126 61.6 240. 26.2 380. 16.2 130 ± ± ± ± ± ± ± 57.5 125 16.7 85 3.334 18.000 160. Hz 16.7 MHz 14.7 MHz 12.6 MHz 10 MHz 50.4 140. 30.000 141 19.00 131 17.000 103 ± ± ± ± ± ± ± ± L80 B70 B70 B70 B70 L80 B69 B69 B69 B69 B70 B70 B70 88.000 8.333 8.333 5.000 61.000 11.667 137.000 10.000 85.000 11.667 21.667 11.667 120.000 16.667 ± ± ± er et al. (1967), G81 - Geldzahler & Witzel (1981), G91 - Grego ± ± ± ± ± ± L80 B70 B70 B70 B70 L80 L80 L80 L80 L80 B70 B70 B70 L80 B69 B69 B69 B69 L80 L80 L80 L80 uliny-Toth (1973), K81 - Kuehr et al. (1981), L80 - Laing & Pea S95 - Slee (1995), V75 - Viner & Erickson (1975), W75 - Wills (1 2 - Colla et al. (1972), C73 - Colla et al. (1973), C83 - Condon e 29.900 7.000 37.950 21.000 120. 11.667 24.000 54.050 19.2 95.00 ± ± ± ± ± B69 B70 101 Table 3 Low frequency radio data 9. 8.000 8.000 64.400 8.000 60.950 7.000 65.550 7.788 31.000 8.000 96.600 6.000 69.000 7.000 7.000 77.000 8.000 10.000 67.850 10.000 18.000 97.750 ± ± ± ± ± ± ± ± ± ± ± ± ± ± L80 L80 V75 42. 25.960 8. 71.300 7.316 34.000 9.558 4.248 4.720 73.000 6.608 60.000 6.844 83.000 6.136 65.000 5.428 60.000 5.664 46.000 9.204 46.000 8.496 69.000 5.782 54.000 7.552 99.000 7.552 57.000 11.092 42.000 10.0 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 L80 V75 L80 L80 L80 L80 42.480 30.680 54.280 56.640 46.020 42.480 55.460 F74 - Fanti et al. (1974), F85 - Ficarra et al. (1985), G67 - Gow . (1995), K69 - Kellermann et al. (1969), K73 - Kellermann & Pa 66), S73 - Shimmins & Wall (1973), SJ95 - Steppe et al. (1995), 1969), B70 - Braude et al. (1970), C70 - Colla et al. (1970), C7 3.019 31.860 5.041 47.200 8.024 64. 3.008 57.820 7.048 75.520 3.008 68.440 4.014 4.011 37.760 3.00 58.9 ± ± ± ± ± ± ± ± ± A68 L80 A68 L80 L80 A68 A68 L80 L80 A68 L80 L80 A68 L80 L80 B70 L80 B70 B70 B70 B70 A68 L80 L80 42.016 46.015 54.013 24.010 34.010 40.120 42.015 53.015 0 33.040 3.79 36.580 2.71 34.0 ± ± ± C04 C07 A68 H95 M05 3.019 ± K81 3C 268.4 20.72 3C 454.0 4C 16.49 3C 432 3C 325 3C 208 3C 245 3C 212 3C 270.1 30.8 3C 318 3C 287 3C 204 26.64 3C 205 3C 191 3C 43 23.004 3C 14 3C 190 3C 186 3C 181 Name3C 9 80MHz 74 MHz 60 MHz 38 MHz 26.3 MHz 25.0MHz 22.25 MHz 20.0 M (1990), W96 - Waldram et al. (1996), C07- Cohen et al. (2007) Hales et al.(1981), (1990), M05 H93 - - Mack Hales et et al. al. (2005), P66 (1993), - H95 Pauliny-Toth - et Hales al. et (19 al (1998), C04 - Cohen et al. (2004), D96 - Douglas et al. (1996), References: A68 - Aslanian et al. (1968), B69 - Braude et al. (