Uncovering the deeply embedded AGN activity in the nuclear regions of the interacting galaxy Arp299
A. Alonso-Herrero, P. F. Roche, P. Esquej, O. Gonzalez-Martin, M. Pereira-Santaella, C. Ramos Almeida, N. A. Levenson, C. Packham, A. Asensio Ramos, R. E. Mason, J. M. Rodriguez Espinosa, C. Alvarez, L. Colina, I. Aretxaga, T. Diaz-Santos, E. Perlman, C. M. Telesco
aa r X i v : . [ a s t r o - ph . C O ] N ov Draft version February 8, 2018
Preprint typeset using L A TEX style emulateapj v. 5/2/11
UNCOVERING THE DEEPLY EMBEDDED AGN ACTIVITY IN THE NUCLEAR REGIONS OF THEINTERACTING GALAXY ARP 299
A. Alonso-Herrero , P. F. Roche , P. Esquej , O. Gonz´alez-Mart´ın , M. Pereira-Santaella , C. RamosAlmeida , N. A. Levenson , C. Packham , A. Asensio Ramos , R. E. Mason , J. M. Rodr´ıguez Espinosa , C.Alvarez , L. Colina , I. Aretxaga , T. D´ıaz-Santos , E. Perlman , and C. M. Telesco Draft version February 8, 2018
ABSTRACTWe present mid-infrared (MIR) 8 − µ m spectroscopy of the nuclear regions of the interactinggalaxy Arp 299 (IC 694+NGC 3690) obtained with CanariCam (CC) on the 10.4 m Gran TelescopioCanarias (GTC). The high angular resolution ( ∼ . − . ′′ ) of the data allows us to probe nu-clear physical scales between 60 and 120 pc, which is a factor of 10 improvement over previous MIRspectroscopic observations of this system. The GTC/CC spectroscopy displays evidence of deeplyembedded Active Galactic Nucleus (AGN) activity in both nuclei. The GTC/CC nuclear spectrum ofNGC 3690/Arp 299-B1 can be explained as emission from AGN-heated dust in a clumpy torus withboth a high covering factor and high extinction along the line of sight. The estimated bolometricluminosity of the AGN in NGC 3690 is 3 . ± . × erg s − . The nuclear GTC/CC spectrum ofIC 694/Arp 299-A shows 11 . µ m polycyclic aromatic hydrocarbon (PAH) emission stemming froma deeply embedded ( A V ∼
24 mag) region of less than 120 pc in size. There is also a continuum-emitting dust component. If associated with the putative AGN in IC 694, we estimate that it wouldbe approximately 5 times less luminous than the AGN in NGC 3690. The presence of dual AGNactivity makes Arp 299 a good example to study such phenomenon in the early coalescence phase ofinteracting galaxies.
Subject headings: galaxies: nuclei — galaxies: Seyfert — infrared: galaxies — galaxies: individual(Arp 299, NGC 3690, IC 694) INTRODUCTION
The interacting galaxy Arp 299 (Mrk 171,IC 694+NGC 3690) was identified as a mid-infrared(MIR) luminous source (Rieke & Low 1972) morethan 40 years ago. Subsequently, Gehrz et al. (1983)detected two bright 10 µ m sources coincident with thegalaxy nuclei (see also Telesco et al. 1985), referred toas Arp 299-A or the nucleus of the eastern component(IC 694) and Arp 299-B or the nucleus of the westerncomponent (NGC 3690). Source B is further resolvedinto B1, the MIR-bright nucleus, and B2, the optical-bright source (see Fig. 1). Two other bright MIR sourcesare in the system overlapping region: Arp 299-C andArp 299-C ′ . The infrared (IR) luminosity of the systemis L IR = 6 . × L ⊙ (for a distance D = 44 Mpc),which puts it in the luminous IR galaxy (LIRG) Instituto de F´ısica de Cantabria, CSIC-UC, 39005 San-tander, Spain; E-mail: [email protected] Augusto Gonz´alez Linares Senior Research Fellow Astrophysics Department, University of Oxford, OxfordOX1 3RH, UK Centro de Astrobiolog´ıa, CSIC-INTA, 28035 Madrid, Spain Instituto de Astrof´ısica de Canarias, 38205 La Laguna,Spain Universidad de la Laguna, 38205 La Laguna, Spain Istituto di Astrofisica e Planetologia Spaziali, INAF, 00133Rome, Italy Gemini Observatory, La Serena, Chile University of Texas at San Antonio, San Antonio, TX 78249 Gemini Observatory, Hilo HI 96720 INAOE, 72000 Puebla, Mexico Spitzer Science Center, Caltech, Pasadena, CA 91125 Florida Institute of Technology, Melbourne, FL 32901 Department of Astronomy, University of Florida,Gainesville, FL 32611 category.Most of the Arp 299 IR luminosity arises from in-tense star formation (SF) activity across the two galax-ies (see Gehrz et al. 1983; Charmandaris et al. 2002;Alonso-Herrero et al. 2000, 2009, AAH00 and AAH09,hereafter). However, there is also evidence of obscuredactive galactic nucleus (AGN) activity in the galaxy nu-clei.Based on hard X-ray observations, the nucleus ofNGC 3690/Arp 299-B1 was long suspected to havea Compton-thick AGN (see e.g. Ballo et al. 2004;Gonz´alez-Mart´ın et al. 2009; Pereira-Santaella et al.2011). Moreover, there is a hot dust continuum asso-ciated with an obscured AGN that contributes substan-tially to the nuclear IR emission (Gallais et al. 2004,AAH00, AAH09). Although initially thought not to bea Seyfert galaxy, Garc´ıa-Mar´ın et al. (2006) measuredSeyfert-like optical line ratios for Arp 299-B1.The case for the presence of an AGN in the nuclearregion of IC 694/Arp 299-A is not as clear. Apart fromthe tentative X-ray evidence (Ballo et al. 2004), per-haps the most convincing argument is the identificationof a flat spectral radio source among the large number ofcompact radio sources detected in the nuclear region ofthis galaxy (Neff et al. 2004; P´erez-Torres et al. 2010).In this paper we present 8 − µ m spectroscopy ofthe two nuclei of Arp 299 obtained with CanariCam(CC, Telesco et al. 2003) on the 10.4 m Gran TelescopioCanarias (GTC). The GTC/CC high-angular resolution0 . − . ′′ ( ∼ −
120 pc for Arp 299 distance) is approx-imately a factor of 10 improvement over previous MIRspectroscopy of this system (Dudley 1999; Gallais et al.2004, AAH09). We use the GTC/CC spectroscopy to Alonso-Herrero et al.
Figure 1.
GTC/CC 8 . µ m acquisition images of IC 694 (left panel) and NGC 3690 (right panel) shown on a linear flux scale. Theprojected separation between the galaxy nuclei IC 694/Arp 299-A and NGC 3690/Arp 299-B1 is approximately 21 ′′ or 4.5 kpc, as canbe seen from figure 5 of AAH09. We also label other IR-emitting sources in NGC 3690: B2, C, and C ′ . The lines indicate the location,orientation, and width of the slits. The images have been smoothed with a Gaussian function with σ = 0 . shed light on the processes giving rise to the deeply nu-clear embedded emission in Arp 299-A and Arp 299-B1. CANARICAM OBSERVATIONS AND DATA REDUCTION
We obtained mid-IR high-angular resolution long-slitspectroscopy of the nuclear regions of Arp 299 using CCon the GTC. The observations are part of our GTC/CCAGN guaranteed time program (PI C. Telesco). We willuse these observations together with an ESO/GTC largeprogram (PI A. Alonso-Herrero) to conduct a MIR sur-vey of approximately 100 local AGN. We used the lowspectral resolution 10 µ m grating, which covers the N -band ∼ . − µ m with nominal R = λ/ ∆ λ ∼ ′′ wide slit was oriented at 345 ◦ so that Arp 299-B1 and C could be observed simultaneously (see Fig. 1,right panel). The plate scale of the CC 320 ×
240 Si:Asdetector is 0.08 ′′ /pixel, which provides a field of view inimaging mode of 26 ′′ × ′′ .The observations were taken in queue mode under pho-tometric conditions using the standard MIR chop-nodtechnique on January 29 and June 25 2013 for Arp 299-A and Arp 299-B1+C, respectively. First, we obtainedan acquisition image (Fig. 1) through the Si-2 filter( λ c = 8 . µ m and ∆ λ cut = 1 . µ m, 50% cut-on/off) to en-sure optimal placement of the slit. The on-source integra-tion times for the spectroscopy were 1011 s for Arp 299-Aand 354 s for Arp 299-B1+C. We also observed standardstars in imaging and spectroscopic mode to provide thephotometric calibration, telluric correction, and slit losscorrection. We used the standard stars and galaxy nucleito measure an angular resolution at 8 . µ m. of 0 . − . ′′ and 0 . ′′ (full width half maximum, FWHM) for Arp 299-A and Arp 299-B1+C, respectively.We reduced the data using the CC pipeline redcan (Gonz´alez-Mart´ın et al. 2013) that includes stacking ofthe individual observations, wavelength calibration, tracedetermination, spectral extraction, and flux calibration.The core emission of the two nuclei appears unresolvedat the resolution of the GTC/CC acquisition images. Weextracted the 1D nuclear spectra in an optimal way fora point source with the aperture size increasing withwavelength to include all the source flux. Finally, weused the observations of the standard stars to correct forslit losses. We estimated a 20% total uncertainty of the Figure 2.
GTC/CC spectra (open symbols) and
Spitzer /IRSspectra (black line) of the nuclear regions of Arp 299-B1 (upperpanel) and Arp 299-A (lower panel) extracted as point sources. Wemark emission lines and PAH features. The noise in the GTC/CCspectra around 9 . µ m is due to the O telluric band. GTC/CC spectra due to flux calibration and correctionfor the point source extraction.To compare with the large scale emission ( ∼ Spitzer (see AAH09 for full details about the observa-tions). We used cubism (Smith et al. 2007) to extracteeply embedded AGN activity in the nuclear regions of Arp 299 3
Table 1 clumpy torus models for NGC 3690/Arp 299-B1 clumpy torus parameter Prior/Range FitTorus width, σ torus ◦ − ◦ ◦ +7 − Torus radial thickness, Y −
100 51 +28 − No. clouds along equator, N −
15 10 +3 − Cloud optical depth, τ V −
150 45 +15 − Cloud radial distribution r − q , q − . +0 . − . Viewing angle, i ◦ − ◦ ◦ +8 − A V (foreground) 16 mag ± +3 − Notes.— In the clumpy models the torus radial thickness isdefined as Y = R o /R d , where R o is the outer radius and R d isthe inner radius. × . ′′ × . ′′ ) Short-Low SL1 ( ∼ − µ m)spectra. We corrected the spectra for point source emis-sion as in Pereira-Santaella et al. (2010). Fig. 2 showsthe GTC/CC nuclear spectra of NGC 3690/Arp 299-B1(top) and IC 694/Arp 299-A (bottom) together with the Spitzer /IRS spectra. NUCLEAR REGION OF NGC 3690/ARP 299-B1
Nuclear versus circumnuclear spectra
The GTC/CC nuclear spectrum of NGC 3690 was ob-served under excellent seeing conditions, FWHM . µ m =0 . ′′ . Thus the unresolved MIR emission originates froma region of ≤
60 pc in size. The nuclear spectrumshows dust continuum emission likely produced by theobscured AGN and a moderately deep 9 . µ m silicate fea-ture. The apparent depth of the silicate feature, definedas S Si = ln( f cont /f feature ), is S Si = 0 .
8. There is faintnuclear 11 . µ m polycyclic aromatic hydrocarbon (PAH)emission (see Fig. 2) with an equivalent width (EW) of < . µ m.The ∼ × Spitzer /IRS spectrum showsa composite (AGN/SF) nature. It shows 8.6 and11 . µ m PAH features together with a strong contin-uum due to hot dust (see also Gallais et al. 2004,AAH09). The EW of the 11 . µ m PAH featuremeasured using a local continuum is ∼ . µ m(Pereira-Santaella et al. 2010), which is typical of AGN(see Hern´an-Caballero & Hatziminaoglou 2011). Thisagrees with the presence of the unresolved IR-brightsource B1, which has an important hot dust contribu-tion at λ ≥ µ m based on the observed near-IR COindex, possibly associated with an AGN. On scales of afew arcseconds, B1 is also surrounded by star clustersand bright H ii regions (AAH00). Modeling with the clumpy torus models
Clumpy torus models reproduce satisfactorily thenuclear IR emission of Seyfert galaxies and givean estimate of the AGN bolometric luminosity(Ramos Almeida et al. 2009, 2011; H¨onig et al. 2010;Alonso-Herrero et al. 2011). In this section we modelthe IR spectral energy distribution (SEDs) and MIRspectrum of NGC 3690 to estimate the AGN’s bolometricluminosity.We constructed the nuclear IR SED of NGC 3690 withthe GTC/CC Si-2 8 . µ m point source measurement of1050 ±
200 mJy and the
Hubble Space Telescope /NICMOS2 . µ m non-stellar measurement (AAH00). We includedthe Keck 1 ′′ -diameter 3.2 and 17 . µ m (Soifer et al. Figure 3.
Best median fit using the clumpy torus models (blackdotted line) and 1 σ confidence (shaded region) to the nuclear IRSED (filled blue dots and arrows) and MIR GTC/CC spectrum(re-sampled to 45 data points, blue line) of NGC 3690/Arp 299-B1. The inset shows the GTC/CC spectrum (filled blue dots witherrors) and best fit models as above. . µ m8.5 ′′ -beam flux densities (Charmandaris et al. 2002) asupper limit since they are probably contaminated by non-AGN emission. Fig. 3 shows the nuclear IR SED.We used the Nenkova et al. (2008) clumpy torus mod-els, also known as clumpy , and the Bayesian fitting rou-tine bayesclumpy (Asensio Ramos & Ramos Almeida2009) to fit the nuclear IR emission of NGC 3690. ABayesian approach allows to handle properly the intrin-sic degeneracies of the clumpy models, provides proba-bility distributions of the fitted parameters, and allowsthe use of priors for the torus parameters.The detection of a nuclear water maser in Arp 299-B1(Tarchi et al. 2011) implies an almost edge-on view ofthe AGN. We restricted the range of the viewing anglesto i = 60 ◦ − ◦ as a prior. Foreground dust is clearlypresent in the nuclear region of this galaxy (see the opti-cal to near-IR color map in figure 9 of AAH00). We setthe prior for the foreground extinction to a Gaussian dis-tribution centered at 16 mag with a 7 mag width, basedon the estimates of AAH00. We use the extinction curveof Chiar & Tielens (2006). For the rest of the clumpy model parameters we used the full range (see Table 1).The best fit clumpy torus model is shown in Fig. 3 andthe parameters are listed in Table 1. The fit is achievedwith a wide torus, a highly inclined view, and a largenumber of clouds along the equatorial direction. Thisresults in a high covering factor of f ∼ . A LOS V (torus) ∼
456 mag domi-nates over the foreground extinction.The best-fit model to the nuclear IR emission pro-vides an estimate of the AGN bolometric luminosity of L bol (AGN) = 3 . ± . × erg s − . This is about afactor of 10 higher than the 0 . −
100 keV intrinsic lu-minosity derived for this system with
BeppoSAX with N H ∼ . × cm − (Della Ceca et al. 2002). Alter-natively we can use the hard X-ray vs. 12 µ m correlationobserved for local Seyfert galaxies (Gandhi et al. 2009;Levenson et al. 2009) and the nuclear 12 µ m monochro-matic luminosity. We predict an intrinsic L −
10 keV =4 . ± . × erg s − . Alonso-Herrero et al.The ratio of observed (Ballo et al. 2004;Pereira-Santaella et al. 2011) to predicted intrinsic2 −
10 keV luminosities is of the order of 500, whichwould only be compatible with the maximum value of N H ∼ × cm − fitted from the BeppoSAX data(Della Ceca et al. 2002). Both MIR estimates of theAGN luminosity are higher than direct measurements,although the highest possible N H value would producecompatible luminosities. A possibility would be that thenuclear MIR emission had an important contributionfrom SF. AAH00 estimated that the SF activity inArp 299-B1 is approximately five times less than inArp 299-A. Then using the nuclear 12 µ m fluxes ofArp 299-A and Arp 299-B1 we predict that SF wouldonly contribute ∼
15% of the observed 12 µ m nuclearemission from Arp 299-B1. We would therefore favorthe high N H value rather than a strong contaminationfrom nuclear SF. Finally, it is possible that the hardX-ray vs. MIR relation might not be applicable to allCompton-thick AGN. NUCLEAR REGION OF IC 694/ARP 299-A
Nuclear versus circumnuclear spectra
The GTC/CC nuclear spectrum of IC 694 (Fig. 2)shows clear PAH emission at 8.6 and 11 . µ m in-dicative of the presence of SF on nuclear scales of ≤
120 pc. This is consistent with results fromhigh angular resolution near-IR (AAH00) and radio(P´erez-Torres et al. 2009) observations of this nucleus.We measured a GTC/CC nuclear EW(11 . µ m PAH) =0 . ± . µ m, which is lower than that measured fromthe kpc-scale Spitzer /IRS spectrum (0 . ± . µ m,see Pereira-Santaella et al. 2010). These are typi-cal of galaxies with a composite (AGN/SF) activity(Hern´an-Caballero & Hatziminaoglou 2011). Moreover,the decreased nuclear EW with respect to the 1 kpc-scaleone also indicates a higher relative contribution in the nu-clear region from continuum emission, possibly producedby dust heated by the putative AGN. This is similarto findings for nearby Seyferts with nuclear SF activity(Esquej et al. 2013). Fit to the nuclear spectrum
Deep silicate features, as in the nuclear spec-trum of Arp 299-A, cannot be reproduced withthe clumpy torus models (Levenson et al. 2007;Alonso-Herrero et al. 2011; Gonz´alez-Mart´ın et al.2013). Instead, Levenson et al. (2007) demonstratedthat spherical dusty shell models where the nuclearsource is deeply embedded in a smooth distribution ofgeometrically and optically thick material can producesuch deep features. Moreover, the reprocessed emissionin these models does not depend on the input spectrumof the heating source. We obtained the best fit shellmodel to the GTC/CC spectrum of Arp299-A with atotal optical depth through the shell of τ V = 162, amaximum temperature at the inside surface of the shellof T max = 1500 K, Y = 100 (defined as in Table 1),and an r − density profile. These are similar to thevalues obtained for the deep silicate feature observedin the ultraluminous IR galaxy IRAS 08572+3915(Levenson et al. 2007). While the overall fit toArp 299-A is reasonable, it fails to reproduce the shape Figure 4.
Fit to the GTC/CC nuclear spectrum ofIC 694/Arp 299-A. The error bars used for the fit are estimatedfrom the scatter of the points in narrow wavelength bins. Theblue dashed line is the best fit with the spherical dusty shell mod-els of Levenson et al. (2007). The dot-dashed red line is the fitusing a black-body + the Orion bar PAH spectrum with silicateabsorption as modelled with the Trapezium silicate grains follow-ing Roche et al. (2007). The solid green line includes additionallya power-law component that fills in the ∼ − µ m spectral range.The error bars around the [Ne ii ]12 . µ m line were increased arti-ficially to exclude it from the fit. of the silicate feature around 10 − µ m (see Fig. 4).Also, the dusty shell model does not fit the strongnuclear PAH emission, as it does not include a starformation component.Alternatively, we fitted the GTC/CC spectrum witha combination of emission and absorption componentsusing the technique described by Roche et al. (2007).We first tried a black-body + the Orion Bar PAH spec-trum suffering silicate absorption. This fit is shown as adashed line in Fig. 4 and has χ /N = 5 .
9, where N isthe number of degrees of freedom. The silicate opticaldepth is τ . µ m = 1 . A V = 21 mag, assuming a dustscreen geometry) with the Trapezium silicate grain pro-file, which is representative of those of molecular clouds.Using the curve of the µ Cephei supergiant gave a poorerfit, but these silicate grains seem to be more appro-priate for sources with extremely deep silicate features(Roche et al. 2007).Although these two components provide quite a goodqualitative agreement over most of the spectrum, the fitfalls below the data points between 9 and 11 µ m, as foundwith the dusty shell model fit above. To improve the fit,we added an IR power-law component ( f ν ∝ ν − , seeRamos Almeida et al. 2009). This fills in the 9 − µ mspectrum (Fig. 4) and improves the fit ( χ /N = 4 . τ . µ m = 1 . A V =24 mag, in agreement with other IR estimates (Gallais etal. 2004, AAH00, AAH09).The power-law component contributes approximately35% of the observed 9 µ m flux density in the nu-clear GTC/CC spectrum of IC 694. We propose thatthis emitting component might be associated with dustheated by the putative AGN and also be responsi-ble for the decreased nuclear EW(11 . µ m PAH). Fromthe power-law monochromatic 12 µ m luminosity we esti-mated an intrinsic L − ∼ ± . × erg s − usingthe Gandhi et al. (2009) relation, as in Section 3.2. Theeeply embedded AGN activity in the nuclear regions of Arp 299 5putative AGN in IC 694 would be about 5 times less lu-minous than that in NGC 3690. DISCUSSION AND CONCLUSIONS
We obtained GTC/CC high-angular resolution (0 . − . ′′ , ∼ −
120 pc) MIR spectra of the interacting galaxyArp 299. We presented evidence of deeply embeddedAGN activity in both nuclei thanks to the improved an-gular resolution of these data of almost a factor of 10 withrespect to previous MIR spectroscopy of this system.The GTC/CC nuclear spectrum ofNGC 3690/Arp 299-B1 shows a strong dust con-tinuum with a moderate silicate feature in absorption.The nuclear 11 . µ m PAH feature is weak comparedwith that in the 1 kpc-scale Spitzer /IRS spectrum.We fitted the IR emission of NGC 3690 with the clumpy models. The dust is located in a highcovering factor torus and heated by an AGN with L bol (AGN) = 3 . ± . × erg s − . The torus highextinction along the line of sight dominates over theforeground extinction.The GTC/CC nuclear (region ≤
120 pc in size) spec-trum of IC 694/Arp 299-A shows a very deep silicate fea-ture reflecting the highly embedded nature of this sourcewith a fitted τ . µ m = 1 . A V = 24 mag, for a dustscreen geometry). There is strong nuclear SF based onthe detection of the 8 . µ m and 11 . µ m PAH features.The lower EW of the nuclear 11 . µ m PAH feature com-pared with the kpc-scale IRS value suggests the presenceof a MIR continuum that also fills in the silicate featurewithin ∼ − µ m. If this nuclear continuum emittingcomponent is due to dust heated by an AGN, then weestimated an AGN luminosity about five times less thanthat in NGC 3690.The interaction of the Arp 299 system started at least750 Myr ago (Hibbard & Yun 1999) but the system hasnot yet fully merged. The projected nuclear separationis ∼ . ∼
15% of the total IR luminos-ity of the system. This relatively small AGN contribu-tion is typical of local LIRGs (see Petric et al. 2011;Alonso-Herrero et al. 2012).Numerical simulations predict that interaction-inducedAGN activity is common. However, dual AGN activ-ity with L bol (AGN) > erg s − would only occur si-multaneously for approximately 10% of the duration ofthe interaction (see e.g., van Wassenhove et al. 2012).Therefore, Arp 299 represents an interesting case studyto test theoretical predictions for dual AGN activity dur-ing the early stages of galaxy interactions.We are extremely grateful to the GTC staff for theirconstant and enthusiastic support. We also thankan anynomous referee for comments that helped im-prove the paper. The following Spanish Plan Na-cional de Astronom´ıa y Astrof´ısica grants are acknowl-edged: AYA2009-05705-E (AAH, PE, CRA, and MPS),AYA2010-21887-C04 (CRA and JMRE), AYA2010-18029(AAR), AYA2010-21161-C02-01 (LC), and AYA2012-39168-C03-01 (JMRE and OGM). AAR also acknowl-edges financial support through the Ram´on y Cajal fel- lowships and Consolider-Ingenio 2010 CSD2009-00038.CP and CMT acknowledge support from NSF grants0904421 and AST-903672, respectively. NAL and RMare supported by the Gemini Observatory, which is op-erated by the Association of Universities for Researchin Astronomy, Inc., on behalf of the international Gem-ini partnership of Argentina, Australia, Brazil, Canada,Chile, and the United States of America.Based on observations made with the Gran Telesco-pio Canarias (GTC), installed in the Spanish Observato-rio del Roque de los Muchachos of the Instituto de As-trof´ısica de Canarias, in the island of La Palma. Basedparty on observations obtained with the Spitzer SpaceObservatory , which is operated by JPL, Caltech, underNASA contract 1407. This research has made use ofthe NASA/IPAC Extragalactic Database (NED) whichis operated by JPL, Caltech, under contract with theNational Aeronautics and Space Administration.
REFERENCESAlonso-Herrero, A., Pereira-Santaella, M., Rieke, G. H., &Rigopoulou, D. 2012, ApJ, 744, 2Alonso-Herrero, A., Ramos Almeida, C., Mason, R., AsensioRamos, A., Roche, P. F. et al. 2011, ApJ, 736, 82Alonso-Herrero, A., Rieke, G. H., Colina, L. et al. 2009, ApJ, 697,660 (AAH09)Alonso-Herrero, A., Rieke, G. H., Rieke, M. J., & Scoville, N. Z.2000, ApJ, 532, 845 (AAH00)Asensio Ramos, A., & Ramos Almeida, C. 2009, ApJ, 696, 2075Ballo, L., Braito, V., Della Ceca, R., Maraschi, L., Tavecchio, F.,& Dadina, M. 2004, ApJ, 600, 634Charmandaris, V., Stacey, G. J., & Gull, G. 2002, ApJ, 571, 282Chiar, J. E., & Tielens, A. G. G. M. 2006, ApJ, 637, 774Della Ceca, R. et al. 2002, ApJ, 581, L9Dudley, C. C. 1999, MNRAS, 307, 553Esquej, P., Alonso-Herrero, A., Gonz´alez-Mart´ın, O., H¨onig, S.,Hern´an-Caballero, A. et al. 2013, ApJ, in press(arXiv:1311.0703)Gallais, P., Charmandaris, V., Le Floc’h, E., Mirabel, I. F.,Sauvage, M., Vigroux, L., & Laurent, O. 2004, A&A, 414, 845Gandhi, P., Horst, H., Smette, A., H¨onig, S., et al. 2009, A&A,502, 457Garc´ıa-Mar´ın, M., Colina, L., Arribas, S., Alonso-Herrero, A., &Mediavilla, E. 2006, ApJ, 650, 850Gehrz, R. D., Sramek, R. A., & Weedman, D. W. 1983, ApJ, 267,551Gonz´alez-Mart´ın, O., Rodr´ıguez Espinosa, J.M., D´ıaz-Santos, T.et al. 2013, A&A, 553, A35Gonz´alez-Mart´ın, O., Masegosa, J., M´arquez, I., & Guainazzi, M.2009, ApJ, 704, 1570Hern´an-Caballero, A., & Hatziminaoglou, E. 2011, MNRAS, 414,500Hibbard, J.E., & Yun, M. S. 1999, AJ, 118, 162H¨onig, S., Kishimoto, M., Gandhi, P. et al. 2010, A&A, 515, A23Houck, J. R. et al. 2004, ApJS, 154, 18Levenson, N. A., Sirocky, M. M., Hao, L., Spoon, H. W. W.,Marshall, J. A., Elitzur, M., & Houck, J. R. 2007, ApJ, 645,L45Levenson, N. A., Radomski, J. T., Packham, C., Mason, R. E.,Schaefer, J. J., & Telesco, C. M. 2009, ApJ, 703, 390Neff, S., Ulvestad, J., S., & Teng, S. H. 2004, ApJ, 611, 186Nenkova, M., Sirocky, M. M., Nikutta, R., Ivezi´c, Z., & Elitzur,M. 2008, ApJ, 685, 160Pereira-Santaella, M., Alonso-Herrero, A., Rieke, G. H., et al.2010, ApJS, 188, 447Pereira-Santaella, M., Alonso-Herrero, A., Santos-Lleo, M. et al.2011, A&A, 535, A93P´erez-Torres, M. A., Alberdi, A., Romero-Ca˜nizares, C., & Bondi,M. 2010, A&A, 519, L5P´erez-Torres, M. A., Romero-Ca˜nizares, C., Alberdi, A., &Polatidis, A. 2009, A&A, 507, L17