Draftversion January25,2016 PreprinttypesetusingLATEXstyleemulateapjv.04/17/13 NUSTAR UNVEILS A HEAVILY OBSCURED LOW-LUMINOSITY ACTIVE GALACTIC NUCLEUS IN THE LUMINOUS INFRARED GALAXY NGC 6286 C. Ricci1,2,*, F. E. Bauer1,2,3,4, E. Treister5,2, C. Romero-Can˜izales1,3, P. Arevalo6, K. Iwasawa7, G. C. Privon5, D. B. Sanders8, K. Schawinski9, D. Stern10, M. Imanishi11,12,13 Draft version January 25, 2016 6 ABSTRACT 1 WereportthedetectionofaheavilyobscuredActiveGalacticNucleus(AGN)intheluminousinfrared 0 galaxy (LIRG) NGC 6286, identified in a 17.5ks NuSTAR observation. The source is in an early 2 merging stage, and was targeted as part of our ongoing NuSTAR campaign observing local luminous n and ultra-luminous infrared galaxies in different merger stages. NGC6286 is clearly detected above a 10keVand, by including the quasi-simultaneousSwift/XRT andarchival XMM-Newton andChandra J data, we find that the source is heavily obscured [N ≃ (0.95−1.32)×1024cm−2], with a column H 1 density consistent with being Compton-thick [CT, log(N /cm−2) ≥ 24]. The AGN in NGC 6286 H 2 has a low absorption-corrected luminosity (L2−10keV ∼ 3−20×1041ergs−1) and contributes .1% to the energetics of the system. Because of its low-luminosity, previous observations carried out in ] the soft X-ray band (< 10keV) and in the infrared did not notice the presence of a buried AGN. A NGC6286 has multi-wavelength characteristics typical of objects with the same infrared luminosity G and in the same merger stage, which might imply that there is a significant population of obscured . low-luminosity AGN in LIRGs that can only be detected by sensitive hard X-ray observations. h Keywords: galaxies: active—X-rays: general—galaxies: interactions—X-rays: galaxies—infrared: p galaxies - o r st 1. INTRODUCTION andtheyarethemajorcontributortotheIRenergyden- [a Luminous [LIR(8−1000µm) = 1011 −1012 L⊙] and 2si0t1y0a).t z ≃ 1− 2 (e.g., Caputi et al. 2007, Goto et al. ultra-luminous(LIR ≥1012L⊙)infraredgalaxies(LIRGs The discovery that most, if not all, U/LIRGs are 1 and ULIRGs, respectively) were first discovered in the triggered by galaxy mergers led to the development of v late sixties (Low & Kleinmann 1968; Kleinmann & Low an evolutionary scenario (Sanders et al. 1988) in which 0 1970). With the advent of the Infrared Astronomical two gas rich disk galaxies collide, triggering an in- 0 Satellite (IRAS, see Sanders & Mirabel 1996 for a re- tense phase of star formation in which they are ob- 8 view),whichdiscoveredalargenumberofU/LIRGs,the served as U/LIRGs. This is then followed by a blowout 5 cosmologicalimportanceoftheseobjectsbecameevident. 0 Although they are relatively rare at low redshift, their phase, during which most of the material enshroud- 1. luminosity function is very steep (Le Floc’h et al. 2005), ing the supermassive black hole (SMBH) is blown away and the system is observed as a luminous red quasar 0 (e.g.,Glikman et al.2015andreferencestherein). When 6 1Instituto de Astrof´ısica, Facultad de F´ısica, Pontificia Uni- most of the dust is removed, the system is eventu- 1 versidadCat´olicadeChile,Casilla306,Santiago22,Chile ally observed as a blue quasar. This model is consis- : 2EMBIGGENAnillo v 3MillenniumInstitute ofAstrophysics,Chile tent with the observed increase of the fraction of ob- i 4SpaceScienceInstitute,4750WalnutStreet,Suite205,Boul- scured sources with redshift up to z ≃ 3 (Treister et al. X der,Colorado80301,USA 2010a; Ueda et al. 2014). The Wide-field Infrared Sur- r 5Universidad de Concepcio´n, Departamento de Astronom´ıa, vey Explorer satellite (WISE) has recently found evi- a Casilla160-C,Concepcio´n, Chile 6InstitutodeF´ısicayAstronom´ıa,FacultaddeCiencias,Uni- dence of a new population of very luminous IR sources versidaddeValpara´ıso,GranBretanaN1111,PlayaAncha,Val- (LIR > 1013L⊙), dubbed Hot Dust Obscured Galax- para´ıso,Chile. ies (Hot DOGs, Wu et al. 2012), which might represent 7ICREA and Institut de Ci`encies del Cosmos, Universitat a short evolutionary phase in the evolution of galaxies, de Barcelona, IEEC-UB, Mart´ı i Franqu`es, 1, 08028 Barcelona, and be related to mergers (e.g., Eisenhardt et al. 2012; Spain 8Institute for Astronomy, 2680 Woodlawn Drive, University Stern et al. 2014; Assef et al. 2015). Numerical simula- ofHawaii,Honolulu,HI96822 tions (e.g., Springel et al. 2005) have also shown that 9Institute for Astronomy, Department of Physics, ETH tidal interactions can drive an inflow of material that Zurich,Wolfgang-Pauli-Strasse27,CH-8093Zurich,Switzerland 10JetPropulsionLaboratory, CaliforniaInstitute ofTechnol- triggers and feeds both accretion onto the SMBH and ogy,Pasadena, CA91109, USA starformation. Therefore,mergersmightplayanimpor- 11Subaru Telescope, 650NorthA’ohoku Place, Hilo,Hawaii, tant role in fuelling the SMBH, as it has been suggested 96720, U.S.A. by the discovery that the fraction of active galactic nu- 12National Astronomical Observatory of Japan, 2-21-1 Os- awa,Mitaka,Tokyo181-8588,Japan clei (AGN) in mergers increases with the AGN luminos- 13Department of Astronomical Science, The Graduate Uni- ity (Treister et al. 2012; Schawinski et al. 2012), span- 8ve5r8s8i,tyJafopraAndvancedStudies(SOKENDAI),Mitaka,Tokyo181- n10in4g1efrrgosm−1<to17%0-8a0t%ain2–t1h0ekmeVostlulummininoosiutsyqoufasLa2r−s1w0it∼h *[email protected] 2 Ricci et al. Ar´evalo et al. 2014, Bauer et al. 2015, Koss et al. 2015, Lansbury et al. 2015, Annuar et al. 2015, Puccetti et al. 2015, Ricci et al. 2015). Previous hard X-ray observa- tionsofU/LIRGshavebeencarriedoutwithBeppoSAX (e.g.,Vignati et al.1999),Suzaku(e.g.,Teng et al.2009) and Swift/BAT (Koss et al. 2013). The recent launch of the Nuclear Spectroscopic Tele- scope Array (NuSTAR, Harrison et al. 2013), the first focussingtelescopeinorbitoperatingatE ≥10keV,has opened a new window in the study of U/LIRGs thanks to its unprecedented characteristics. The first studies of the hard X-ray emission of local ULIRGs carried out withNuSTARhaverecentlybeenreportedbyTeng et al. (2015) and Ptak et al.(2015), who show the importance ofsensitivehardX-rayspectratowellconstraintheline- of-sight column density. We report here on the first results of a series of NuS- Figure 1. OpticalimageoftheinteractingpairNGC6286(bot- TAR observations awardedto our groupduring AO-1 as tom)/NGC6285 (top) obtained with the Schulman 32-inch Tele- apartofacampaignaimedatobservingtenlocalLIRGs scopeoftheMountLemmonSkyCenter. ImagecourtesyofAdam indifferentmergerstages(PI:F.E.Bauer). Thesources Block(MountLemmonSkyCenter/UniversityofArizona). wereselectedfromtheGreatObservatoriesAll-skyLIRG Survey (GOALS15, Armus et al. 2009). GOALS is a lo- L2−10 ∼1046ergs−1. cal(z <0.088)samplewhichcontains181LIRGsand21 The contribution of AGN to the overall luminosity of ULIRGs selected from the IRAS Revised Bright Galaxy U/LIRGs has been shown to increase with the IR lu- Sample (Sanders et al. 2003). minosity of the system (e.g., Veilleux et al. 1995, 1999; This paper reportsthe detection of a heavily obscured Imanishi 2009; Imanishi et al. 2010a,b; Nardini et al. AGN in NGC6286 (also referred to as NGC6286S), a 2010; Alonso-Herreroet al. 2012; Ichikawa et al. 2014). LIRG (logLIR/L⊙=11.36, Howell et al. 2010) located Due to the great opacity of the nuclear region, a clear at z=0.018349 (i.e., a luminosity distance of d = L identification of AGN in U/LIRGs is often complicated. 76.1Mpc), which was not previously detected above Mid-IR (MIR) properties have been used to estimate 10keV (Koss et al. 2013). The source has a star- the relative contribution of accretion onto the SMBH formation rate (SFR) of 41.3M⊙yr−1 (Howell et al. and star formation to the bolometric luminosity. This 2010), is in an early merging stage (i.e., stage B or 2, has been done by exploiting 5–8µm spectroscopy (e.g., following the classification of Stierwalt et al. 2013), and Nardini et al. 2010), and the characteristics of several is interacting with the galaxy NGC6285 (NGC6286N), features in the L (3–4µm) and M (4–5µm) bands locatedatadistanceof∼1.5arcmin(∼33kpc,projected (Imanishi & Dudley2000;Risaliti et al.2006;Sani et al. distance, see Fig.1 and panel four of Fig.2). The source 2008; Risaliti et al. 2010): the 3.3µm polycyclic aro- is also known to host a OH megamaser (Baan et al. matic hydrocarbon (PAH) emission feature, the bare 1998). TheonlypreviousX-raystudyofthissource,car- carbonaceous 3.4µm absorption feature, and the slope ried out using XMM-Newton observations, did not find of the continuum. The 6.2µm (e.g., Stierwalt et al. any evidence of an AGN (Brightman & Nandra 2011). 2013,2014)and7.7µmPAHfeatures(e.g.,Veilleux et al. The XMM-Newton can in fact be well represented by 2009), the presence of high-excitation MIR lines (e.g., a model taking into account only a collisionally-ionized [NeV]14.32µm, Veilleux et al. 2009), or the radio prop- plasmaandanunabsorbedpower-lawcomponent,repre- erties (e.g., Parra et al. 2010, Romero-Can˜izales et al. senting thermal emission from the starburst and X-ray 2012a, Vardoulaki et al. 2015) have also been used to radiation produced by X-ray binaries, respectively. Pos- infer the presence of a buried AGN. sibleevidence ofveryfaintAGN activityhasbeen found X-ray observations are a very powerful tool to detect studying the near-IR to radio spectral energy distribu- accretingSMBHsandtodisentanglethecontributionsof tion (SED) (Vega et al. 2008), and could be inferred by star formation and AGN emission to the total luminos- the detection of [NeV] lines, although the detection of ity of U/LIRGs. Studies performed so far using XMM- thesefeatureshasbeenquestionedbyInami et al.(2013), Newton (e.g., Franceschini et al. 2003; Imanishi et al. and due to their weakness they might also be produced 2003; Pereira-Santaellaet al. 2011) and Chandra (e.g., by a young starburst. Ptak et al. 2003; Teng et al. 2005; Iwasawa et al. 2011) The paper is structured as follows. In §2 and §3 we have characterized the properties of a significant num- present the X-ray and radio data used and describe the ber of these systems. However, a significant frac- data reductionprocedures,in§4 we reportonthe X-ray tion of U/LIRGs might be heavily obscured (e.g., spectral analysis of NGC6286, in §5 we discuss our re- Treister et al. 2010b; Bauer et al. 2010), and X-rays at sults by taking into account the multi-wavelength prop- energies . 10keV are strongly attenuated in Compton- ertiesofNGC6286,andin§6wesummarisethemainre- thick (CT, NH ≥ 1024cm−2) AGN. Observations car- sults of our work. Throughoutthe paper we adoptstan- ried out in the hard X-ray band (≥ 10keV) are less dard cosmological parameters (H = 70kms−1Mpc−1, 0 affected by absorption, and can be used to probe Ω =0.3, Ω =0.7). m Λ nuclear X-ray emission even in highly obscured sys- tems (e.g., Balokovi´cet al. 2014, Gandhi et al. 2014, 15 http://goals.ipac.caltech.edu NuSTAR unveils a low-luminosity heavily obscured AGN in the LIRG NGC 6286 3 1) Chandra (0.2-2 keV) A B Core C 4’’ 3.95513" 0 0.6 1.2 1.8 2.4 3 3.6 4.2 4.8 5.4 6 3) NuSTAR (3-24 keV) 4) Spitzer/IRAC (3.6 micron) NGC 6285 NGC 6285 NGC 6286 A NGC 6286 B Core C 1’ 22’’ 1.00002’ 22.1632" 0.05 0.084 0.12 0.15 0.18 0.22 0.25 0.29 0.32 0.35 0.39 0.1 0.2 0.4 0.9 1.7 3.4 6.9 13.7 27.5 54.7 109.0 Figure 2. ChandraACIS-S(panelone,0.2–2keV;paneltwo,3–8keV),NuSTARFPMA(panelthree,3–24keV)andSpitzerIRAC(panel four, 3.6µm) images of NGC6286. The Chandra 0.2–2keV image shows a clear extended structure of ∼ 12arcsec size (∼ 4.4kpc). The fourregionsshowninpanelsoneandfourrepresentthe3–8keVcore,thenorth,central,andsouthregionsdiscussedin§4.1. The1.4GHz VLA FIRST radiocontour is illustratedinpanel two, together withthe two radiosources (R1 and R2) detected by our analysis of EVN observationsat5GHz(see§3),whichshowthattheradioemissioncoincideswiththehardX-raycore,suggestingthepresenceofaburied AGN.Theblackcircleandthebluedashedannulus inpanel threecorrespondtotheNuSTARsourceandbackground extraction regions, respectively. The image in panel four was smoothed with a Gaussian kernel of radius 5 pixels. The blue dashed circle in panel four representsthesourceregionusedforXMM-NewtonEPIC/PN. 2. X-RAYOBSERVATIONSANDDATAREDUCTION exposures of 20.8 and 8.9ks. Both PN (Stru¨der et al. 2.1. NuSTAR 2001) and MOS (Turner et al. 2001) data were analysed by reducing first the observation data files (ODFs) us- NuSTAR observed NGC6286 on UT 2015 May 29 ingtheXMM-NewtonStandardAnalysisSoftware(SAS) for 17.5ks. We processed the data using the NuS- version12.0.1(Gabriel et al.2004),andthentherawPN TAR Data Analysis Software nustardasv1.4.1 within and MOS data files using the epchain and emchain HEASOFTv6.16, adopting the latest calibration files tasks, respectively. In order to filter the observations (Madsen et al. 2015). The source is clearly detected in for periods of high background activity we analyzed the the 3–24keV image (panel three of Fig.2). For both EPIC/PNandMOSbackgroundlightcurvesinthe10–12 focal plane modules (FPMA and FPMB) we extracted keVbandandabove10keV,respectively,andfoundthat sourceandbackgroundspectraandlight-curveswiththe bothobservationsshowasignificantbackgroundcontam- nuproducts task. A circular region of 45arcsec was ination. Observation 0203391201 was not used because used for the source16, while the background was ex- the background flux dominates the whole observation tracted from an annulus centred on the X-ray source, (withanaveragecount-rateof6cts−1andaminimumof with inner and outer radii of 90 and 150 arcsec, re- 2cts−1). Observation 0203390701 showed less contam- spectively. The 3–10 and 10–50keV light-curves of the ination, and we filtered the periods of high background sources do not show any evidence of flux variability. activity using a threshold of 2cts−1 for both PN and MOS, which resulted in net exposure times of 2.3 and 2.2. XMM-Newton 4.7ks, respectively. For both cameras we extracted the Two XMM-Newton (Jansen et al. 2001) observations spectrumofthesourceusingacircularregionof20arcsec of NGC6286 (ID: 0203390701 and 0203391201; PI: radius,whilethebackgroundwasextractedfromacircu- Maiolino) were carried out on UT 2005 May 7, with lar region of 40arcsec radius, located on the same CCD of the source and in a zone devoid of other sources. No 16 In the 3–24keV band for photon indices Γ = 0.6−1.8 this significantfluxvariabilityisfoundinthe0.3–10keVband aperture encloses ∼ 65% of the full PSF energy (Lansburyetal. during the XMM-Newton observation. 2015). 4 Ricci et al. archive and applied the pipelined calibration available. InFigure3we showthe contourmapobtainedusing the cleaning algorithm within the Caltech program difmap (Shepherd et al. 1995). No proper flux density could be obtainedwithsuchasmallarraysincethemeasurements arestillsubjectofinstrumentalamplitudeerrors. Wecan however rely on the source structure, as there is enough information to determine phase closure. We find two milli-arcsec sources with S/N > 5, one (R1) at RA= 16h58m31.7374s, DEC = +58◦56′14.705′.′, and the other (R2) at RA=16h58m31.6572s, DEC= +58◦56′14′.′167. These two sources are consistent with the 3–8keV core region (see panel two of Fig.2 and §4.1). We have also extracted and analysed the very long baseline array (VLBA) experiment BC196 observed at 8GHz on UT 2012 January 12. We used the NRAO AstronomicalImage ProcessingSystem (aips) to reduce the data, following standard procedures. We note that Figure 3. EVN(Effelsberg,WesterborkandLovellantennas)con- the source chosen as phase reference (J1651+5805) is tour map of NGC6286 at 5GHz from UT 2005 June 13, imaged with a convolving beam of 8.39×25.94arcsec at 33.71◦ (North not detected in this experiment, and constraints for it throughEast)usingnaturalweighting. arealsonotavailablein the VLBACalibratorsearchen- 2.3. Swift/XRT gine at NRAO. We have resorted to the use of another nearby calibrator (J1656+6012,at 2.22◦) which was ob- TheX-rayTelescope(XRT,0.2–10keV;Burrows et al. served 2min before the NGC6286 scan. We found that 2005) on board Swift (Gehrels et al. 2004) observed there are no sources detected above ∼ 0.8mJy/beam NGC6286 quasi-simultaneously with NuSTAR on UT (3×r.m.s.) in the VLBA observations convolved with 2015 May 29 for 2ks. XRT data were reduced using a3.16×0.94arcsecat29.31◦ beam. Ifanyofthe sources the xrtpipeline v0.13.0 within HEASOFTv6.16. detected with the EVN is the AGN core, we would ex- pectasimilarpeakintensitymeasuredinabaselinewith 2.4. Chandra comparable length as that from Ef-Jb or Ef-Wb base- A Chandra (Weisskopf et al. 2000) ACIS-S lines. The fact that we do not detect any source in the (Garmire et al. 2003) observation of NGC6286 was VLBA observations leaves two possible explanations: i) carried out on UT 2009 September 18 (PI: Swartz) the sources are variable and the VLBA observations are with an exposure of 14.2ks. The data reduction was not sensitive enough; ii) the sources are resolved at res- performed following the standard procedure, using olutions better than ∼ 3arcsec. Although the VLBA CIAO v.4.6. The data were reprocessed using chan- arrayincludesthreetimes asmuchantennasasthe EVN dra repro, and then the spectra were extracted using small array, it also observed the target for only 1/3 of the specextract tool. the time with respect to the EVN, and using antennas The 0.2–2keV Chandra image shows clear evidence of ∼3–100timessmallerthanthoseinthesmallEVNarray. extendedemission(paneloneofFig.2). The3–8keVim- We made the exercise of producing an image with simi- age (panel two) shows instead only a point-like source, laruv-rangeforbothEVNandVLBAobservations. The which does not appear in the 0.2–2keV image. This obtaineduv-coveragesresultintheVLBAbeingsensitive source is located at the center of the galaxy (see panel toemissionclosetoperpendiculartothestructurewede- three) and couldbe associatedwith AGN emission. The tectedwiththeEVN(ataninclination∼50◦),andsince spectra of these different regions are discussed in §4.1. there is no emission in such orientation, the VLBA can- Inordertobeconsistentwiththespectralextractionof not detect any structure, unlike the EVN. Threfore, in XMM-NewtonandSwift/XRT,whichhaveamuchlower ordertobetterconstraintheradioemissionofNGC6286, spatialresolutionthanChandra,theACIS-Ssourcespec- furtherVLBIobservationscoveringproperhouranglesat trumusedforthebroad-bandX-rayfittingwasextracted high sensitivity are needed. from a circular region of 10.5arcsec radius. The back- 4. X-RAYSPECTRALANALYSIS groundspectrum was extractedfrom a circularregionof the same size on the same CCD, where no other source The X-ray spectral analysis was performed within was detected. xspecv.12.8.2 (Arnaud 1996). Galactic absorption in the direction of the source (NG = 1.8 × 1020cm−2, 3. RADIOOBSERVATIONSANDDATAREDUCTION Kalberla et al. 2005) was taken Hinto account by adding Parra et al. (2010) reported VLA observations of photoelectricabsorption(TBabsinxspec, Wilms et al. NGC6286 at 4.8GHz, showing a compact morphology 2000). Abundances were set to the solar value. Spectra with a size of 0.25× 0.21arcsec and a flux density of wererebinned to haveat least20counts per bin in order 15.24mJy. They also observed this galaxy with three of to use χ2 statistics, unless reported otherwise. the most sensitive antennas (Effelsberg, Westerbork and In the following we first present the X-ray spectral Lovell)oftheEuropeanverylongbaselineinterferometry analysis of the extended and nuclear emission revealed (VLBI)Network(EVN)at5GHz,anddetectedfringesin by Chandra (§4.1) and then discuss the spatially inte- eachone of the baselines with amplitudes between 5.36– grated broad-bandX-ray emission (§4.2) considering all 6.11mJy. Wehaveextractedtheseobservationsfromthe observations available. NuSTAR unveils a low-luminosity heavily obscured AGN in the LIRG NGC 6286 5 Figure 4. Left panel: XMM-Newton EPIC/PN, Chandra ACIS-S (using a 10.5 arcsec extraction radius) and NuSTAR FPMA/FPMB spectra of NGC6286. The continuous lines represent the model used in Brightman&Nandra (2011) (apec+zpowerlaw) to reproduce the0.3–10keVspectrumofthesource. Rightpanel: ratiobetweenthedataandthemodel. Theplotsclearlyshowtheimportanceofhard X-raycoveragetofullyunderstandthecharacteristicsoftheX-rayemission. 4.1. Extended and nuclear emission a rather poor fit (C-stat/DOF=48.0/34). This can be improved by adding photoelectric absorption The diffuse soft X-ray emission detected by Chandra (tbabs ×ztbabs×apec, C-stat/DOF=43.7/33),and has an angular size of ∼ 12arcsec, which at the dis- Gal would be consistent with the presence of larger absorp- tanceofthesourcecorrespondsto∼4.4kpc. Thisdiffuse tion in the central part of the edge-on galaxy with emission might either be related to thermal plasma in a respect to other regions. The shock plasma model star-forming region, to X-ray binaries, or to shocks cre- fails to reproduce well the spectrum both without (C- ated by the interaction between outflows from the AGN stat/DOF=48.7/33) and with (C-stat/DOF=43.6/32) and the galactic medium. To analyse the diffuse and an absorption component. nuclear emissionwe extractedthe spectra of the four re- gions shown in panel one of Fig.2. Besides the 3–8keV Region C. A thermal plasma model with kT = core,inordertostudyhowtheextendedemissionvaries, 1.23+0.37keV yields a good fit (C-stat/DOF=34/34), −0.35 we arbitrarilyselected three regions(A, B andC) where while a shock plasma model cannot reproduce well mostofthe 0.2-2keVphotonsweredetected. Due to the the data (C-stat/DOF=45.1/33), and results in kT = s low number of counts we rebinned the spectra to have 1.09+0.73keV and τ ≤1.4×108scm−3. −0.38 u at least 1 count per bin, and used Cash statistics (Cash 1979) to fit the data. In the following we discuss the 4.2. Spatially integrated X-ray emission spectralpropertiesoftheCore,andtheregionsA,Band The XMM-Newton EPIC spectrum of NGC6286 was C. analysed by Brightman & Nandra (2011), who found Core. The spectrum of the core was extracted from that it could be well represented by an unabsorbed a circular region of radius 1.5arcsec centred on the power-law continuum plus a thermal plasma17, with the peak of the 3–8keV emission. Ignoring the data be- photon index fixed to Γ = 1.9. This is in disagree- low 1.2keV to avoid contamination from the diffuse ment with the hard (Γ = −0.17+1.01, see §4.1) 1–8keV soft X-ray emission, and fitting with a power-law model −1.03 spectrum of the 3–8keV core. Fitting the NuSTAR (tbabs ×zpowerlaw in xspec) we obtain a pho- Gal FPMA/FPMB data with a simple power-law model we dtoicnatiinvdeeoxfohfeaΓvy=ab−so0r.p1t7io+−n11..00i13n.thTehniusclloeawr rveagluioen.isFiint-- also find a very flat continuum (Γ=0.49+−00..4461). The low values of the photon index obtained in the 3–8keV and ting the X-ray spectrum using the whole energy range 3-30keVbands couldindicate that the X-rayemissionis with a model that includes also a collisionally ionized highly absorbed. plasma model (tbabs ×zpowerlaw+apec) we ob- Gal While the modelused by Brightman & Nandra(2011) tain C-stat/DOF=20.9/21, Γ ≤ −0.03 and a plasma can reproduce well the XMM-Newton and the spatially temperature of kT = 0.99+−00..2385keV. The 3–8keV core integrated Chandra spectra, it severely under predicts coincides with the 1.4GHz radio emission measured by thehardX-rayfluxinferredbyNuSTAR,asillustratedin the VLA FIRST survey (Becker et al. 1995). Fig.4. Thismightberelatedeithertoheavyobscuration of the X-ray source or to flux variability between XMM- Region A. Fitting the spectrum with a collisionally Newton and NuSTAR observations, although variability ionized thermal plasma model (tbabs ×apec) re- Gal would not be able to explain the very flat hard X-ray sults in a good fit (C-stat/DOF=31.0/37), with kT = spectrum. The Swift/XRT observationallows us to con- 0.91+0.10keV. Applying a spectral model which repro- −0.18 strainthefluxlevelbelow10keVbandatthetimeofthe duces a non-equilibrium plasma created in a shock NuSTARobservations. WefindthattheSwift/XRT0.3– (pshock in xspec) yields C-stat/DOF=31.4/36, a 2and2–5keVfluxes areconsistentwiththat inferredby plasma temperature of kT = 0.88+0.11keV and an up- S −0.13 Chandra and XMM-Newton EPIC/PN observations (see per limit on the ionization timescale of τu ≥ 1.6 × Table1), which implies the lack of significant variability 1012scm−3. Region B. Using the thermal plasma model yields 17 zpowerlaw+apecinxspec 6 Ricci et al. Figure 5. Left panel: unfolded broad-band X-rayspectrum of NGC6286. The black continuous linerepresents the best fit to the data, whilethedot-dashedlineisthethermalplasma,thedashedlineisthescatteredemissionandthedot-dot-dashedlineisthespheremodel. Right panel: ratiobetween thedataandthemodelshownintherightpanel. fuse emission, and allow them to vary only within their Table 1 ObservedX-rayfluxes 90% uncertainties. 4.2.1. PEXRAV Facility Flux To infer the value of the line-of-sight column density (N ) we fitted the joint Swift/XRT, Chandra ACIS- 0.3–2keV 2–5keV H S, XMM-Newton EPIC/PN and MOS, and NuSTAR [10−14ergs−1cm−2] [10−14ergs−1cm−2] FPMA and FPMB data with a model that consists Swift/XRT 8.4+−42..19 2.0+−01..80 of: a) an absorbed power-law with a photon index Chandra 8.8+−00..66 2.0+−00..34 fiofxeAdGtNo(Γe.g.=, N1a.n9d,rcao&nsiPstoeunntdsw1it9h94tah;ePaicvoenrcaeglelivetalaule. XMM-Newton 9.5+−01..81 2.2+−00..24 2005;Ricci et al.2011),b)unabsorbedreprocessedX-ray emissionfromaslab,c)aGaussiantoreproducethe flu- orescentFeKαemissionline (withthe rest-frameenergy fixed to E = 6.4keV), d) a second power-law to re- between the different observations. To further test the Kα produce the scattered component, and e) emission from variability scenario we fitted NuSTAR and the spatially a collisionallyionizedplasma. To reproducethe effect of integrated Chandra spectra with a model that consists obscuration we included both Compton scattering and ofapower-lawplusathermalplasma[tbabs ×(apec Gal photoelectric absorption. Reprocessed X-ray emission + power law)],allowingfordifferentnormalisationsof (excludingfluorescentlines)wastakenintoaccountusing the power-law continuum to vary (fixing Γ = 1.9). We the pexrav model (Magdziarz & Zdziarski 1995). The found that the model cannot reproduce well the spectra (χ2/DOF=28.9/20), with the fit18 showing clear resid- fractionofscatteredflux(fscatt)iscalculatedastheratio betweenthenormalizationat1keVoftheprimarypower uals between 10 and 30keV. This rules out variability law(n )andnscatt. ThewidthoftheGaussianlinewas as a likely explanation for the large ratio between the po po 10–50keV and 2–10keV fluxes. fixed to σ = 40eV, consistent with the results obtained In the following we report the results obtained by byChandra/HETGstudies(e.g.,Shu et al.2010). AnFe adopting several different X-ray spectral models to in- Kαlineat6.4keVisusuallyfoundintheX-rayspectrum fer the properties of the AGN in NGC6286. In order of AGN (e.g., Nandra & Pounds 1994b, Shu et al. 2010, to reduce the possible degeneracies in the models, we Ricci et al. 2014b), and is believed to originate in the constrained the average properties of the diffuse soft X- materialsurroundingtheSMBH(e.g.,Ricci et al.2014a, ray emission. To do this we first extracted the Chan- Gandhi et al.2015andreferencestherein). Inxspecour dra X-ray spectrum of the diffuse emission by exclud- model is: ing from the circular region of 10.5arcsec a circle of tbabs (ztbabs×cabs×zpowerlaw + pexrav + Gal 1.5arcsec centred on the 3–8keV core. We then fitted zgauss + apec + zpowerlaw). the spectrum with a model that includes i) a thermal The model yields a good fit (χ2/DOF=47.2/44) and plasma and ii) a power-lawcomponent (Γ=1.9)to take results in a column density consistent with border- into account the scattered emission. We obtained a nor- line Compton-thick obscuration (N = 1.32+0.82 × malization of the power law nscatt = (1.03 ± 0.37) × H −0.54 po 1024cm−2). Due to the low signal-to-noise ratio, the Fe 10−5phkeV−1cm−2s−1,andatemperatureandnormal- Kα is not spectrally resolved, and only an upper limit izationofthe thermal plasmaof kT =0.77+0.07keVand −0.08 of its equivalent width was obtained (EW ≤ 2318eV), napec =(2.02±0.33)×10−5phkeV−1cm−2s−1, respec- which is consistent with heavy obscuration. tively. In all the spectral models reported below we set nscatt, kT and n to the values obtained for the dif- 4.2.2. TORUS po apec To further study the absorbing material we used the 18 theratioofthepower-lawnormalisationsis≃4. torusmodeldevelopedbyBrightman & Nandra(2011), NuSTAR unveils a low-luminosity heavily obscured AGN in the LIRG NGC 6286 7 values; ii) fixing the inclination angle of mytorusL and mytorusS to θ (S,L) = 0◦, and that of mytorusZ to i θ (Z) = 90◦; iii) adding a second scattered component i with θ (S,L) = 90◦; iv) leaving the normalizations of i the transmitted and scattered component (n and po n ) free to vary. To this model we added a scattered refl component and thermal emission. In xspec the model is: tbabs ×{mytorusZ(90◦)× zpower- Gal law + mytorusS(0◦) + mytorusS(90◦) + gsmooth[mytorusL(0◦) + mytorusL(90◦)] + apec + zpowerlaw}. Due to the low signal-to-noise ratio of the spectrum we couldnot constrainthe differentvalues of NT(S,L) and H Figure 6. Valueof∆χ2=χ2−χ2best(whereχ2bestistheminimum NHT(Z), so their values were tied. The same was done value of the χ2) versus the column density for the different X- for the normalizations of the scattered and transmitted rayspectral models discussedin§4.2. Thehorizontal dashed line components,whilethephotonindexwasleftfreetovary. represents ∆χ2 = 2.7. The plot shows that NGC6286 is heavily This model also yields a goodfit andresults in a line-of- obscured, with NH consistent with the source being CT for the sight column density consistent with heavy obscuration fivemodelsconsidered. (N =8.8+5.1×1023cm−2). which considers reprocessed and absorbed X-ray emis- H −3.8 sion from a spherical-toroidal structure. In this model Theparametersobtainedfromthespectralanalysisare the line-of-sightcolumndensityisindependent ofthe in- reportedinTable3, while inFigure6we showthe values clination angle, which we fixed to the maximum value of ∆χ2 versus N for the models described above. De- permitted (θ = 87.1◦). Similarly to what was done for H i pendingontheX-rayspectralmodeladopted,theintrin- pexrav, we added to the model a power law, to take sic(i.e. absorptionandk-corrected)2–10keVluminosity into account the scattered emission, and a collisionally of NGC6286 is 3−20×1041ergs−1. ionized plasma model. In xspec, the model is tbabsGal(atable{torus1006.fits} + apec + zpow- 5. DISCUSSION erlaw). The X-ray spectral analysis of NGC6286 reported We fixed Γ = 1.9 and tested several values of the half- above clearly shows that the accreting SMBH is heav- opening angle of the torus (θOA = 40◦,60◦,80◦). The ily obscured, possibly by CT material (see Fig.6). The three models are statistically indistinguishable, and in veryflatcontinuumfoundbybothChandra(forthehard all cases we obtained good fits. For the three values X-ray core) and NuSTAR, together with the fact that of θOA the column densities are consistent within the the 1.4GHz emission coincides with the 3–8keV Chan- uncertainties with CT absorption. drapoint-source(Panel2ofFig.2)confirmsthepresence of a heavily obscured AGN. While the buried AGN in 4.2.3. SPHERE NGC6286could be easily identified at hard X-rays,sev- To test the scenario in which the X-ray source is fully eral other diagnostics failed to detect it because of its coveredbytheobscuringmaterialweappliedthesphere low-luminosity. In§5.1weillustratethemostcommonly model (Brightman & Nandra 2011), using the same set- adoptedtechniquestodetectAGNinU/LIRGs,anddis- ting as for the torus model: cuss the case of NGC6286 by exploiting the wealth of multi-wavelength data available for the GOALS sample. tbabs (atable{sphere0708.fits} + apec + Gal In §5.2 we estimate the contribution of the AGN to the zpowerlaw). luminosityofNGC6286,whilein§5.3wediscusstheop- This model provides a good fit (Fig.5), and confirms ticalandradiopropertiesofthe galaxy,comparingthem the presence of heavy obscuration (N = 9.8+4.6 × to those of other similar LIRGs. Finally, in §5.4, we H −3.5 1023cm−2). discuss the presence of heavily obscured low-luminosity AGN in LIRGs. 4.2.4. MYTORUS Next we applied the MYTorus model 5.1. IR and X-ray tracers of AGN activity in U/LIRGs (Murphy & Yaqoob 2009), which considers absorbed AGN in U/LIRGs can be identified in the IR by and reprocessed X-ray emission from a smooth torus several means: i) with the detection of high-excitation with θOA = 60◦, and can be used for spectral fitting as MIR emission lines (e.g., Sturm et al. 2002), and a combination of three additive and exponential table in particular of [NeV]14.32µm and [NeV]24.32µm models: the zeroth-order continuum (mytorusZ), the (e.g., Weedman et al. 2005, Satyapal et al. 2008, scattered continuum (mytorusS) and a component Goulding & Alexander 2009); ii) using the ra- containing the fluorescent emission lines (mytorusL). tios of high-to-low ionization fine-structure emis- We used the decoupled version of MYTorus (Yaqoob sion lines (e.g., [NeV]14.32µm/[NeII]12.8µm and 2012). This was done by: i) allowing the values of [OIV]25.89µm/[NeII]12.8µm; e.g., Lutz et al. 1999, the column density of the absorbing [NT(Z)] and Petric et al. 2011); iii) with the EW of the PAH H reprocessing [NT(S,L)] material to have different features, which tend to be lower in the presence of a H 8 Ricci et al. Table 2 ListofIRandX-raytracersofAGNactivitycommonlyusedforU/LIRGs. (1) (2) (3) (4) (5) (6) Indicator NGC6286 Reference MeanGOALS Threshold AGN [NeV]14.32µm [10−17Wm−2] 0.33±0.11 Dudiketal.(2009) 2.27A ··· ? [NeV]24.32µm [10−17Wm−2] 0.99±0.20 Dudiketal.(2009) ··· ··· ? [NeV]/[NeII] 0.02 Dudiketal.(2009) 0.07B ≥0.1C ✕ [OIV]/[NeII] 0.05 Dudiketal.(2009) 0.03D/0.24E ≥1.75F ✕ Γ2.5−5µm −0.1 Imanishietal.(2010b) ··· ≥1G ✕ Fν(30µm)/Fν(15µm) 5.97 Stierwaltetal.(2013) 8+−21.5H ··· ✕ EW(PAH3.3µm)[nm] 48 Imanishietal.(2010b) ··· <40G ✕ EW(PAH6.2µm)[µm] 0.59 Stierwaltetal.(2013) 0.55H ≤0.3C ✕ τ3.1µm (3.1µm H2Oice) ND Imanishietal.(2010b) ··· >0.3G ✕ τ3.4µm (3.4µm barecarbonaceous) ND Imanishietal.(2010b) ··· >0.2G ✕ τ9.7µm −0.40 Stierwaltetal.(2013) −0.35H ··· ✕ Chandra hardnessratio −0.85±0.07 Thiswork −0.56I >−0.3J ✕ ObservedlogL2−10 [ergs−1] 40.80 Thiswork 41.3K >42L ✕ Radio/FIR fluxratio(q) 2 Uetal.(2012) 2.41±0.29M <1.64N ✕ Thetablelists(1)theindicatorused,(2)thevalueand(3)referenceforNGC6286,(4)themeanvalueforthe GOALSsample,(5)thethresholdusedtoinferthepresenceofAGN,and(6)whetheranAGNwasfoundornot. Notes. ND:notdetected; A medianofthe43detections (18% ofthesample)fromPetricetal.(2011); B median(Petricetal.2011);C thresholdusedbyInamietal.(2013)toestablishasignificantcontribution oftheAGNtotheMIRflux. D medianandE meanfromPetricetal.(2011);F ValueindicatingiftheAGN contributes tomorethan50%oftheMIRflux(Petricetal.2011);G valueusedbyImanishietal.(2010b) H meanvalueforobjectsinthesamemergerstage(B)asNGC6286(Stierwaltetal.2013),the30µm/15µm fluxdensityratioofNGC6286isonlymarginallylowerthantheaveragevaluefortheBmergerstage, and hasavalueconsistentwith63%ofGOALSLIRGs;I medianofthe44objectsreportedinIwasawaetal.(2011); J valueusedtoestablishthepresenceofanAGNandK medianvalue(Iwasawaetal.2011);L Values commonlyusedtoseparateAGNfromstarburstsinthe2–10keVband(e.g.,Szokolyetal.2004, Kartaltepeetal.2010);M meanobtainedforthe64objectsstudiedbyUetal.(2012);N thresholdfor radio-excessdefinedbyYunetal.(2001). bright AGN, since it can destroy PAH molecules (e.g., detected at ∼24.37µm (Inami, private communication). Imanishi et al. 2010b); iv) studying the slope of the Spitzer/IRAC selection provides another important tool 2.5–5µm continuum (Γ2.5−5, e.g., Imanishi et al. 2010b) for identifying AGN (e.g., Lacy et al. 2004, Stern et al. or the continuum 30µm/15µm flux density ratio (e.g., 2005). Using the AGN selection criteria proposed by Stierwalt et al. 2013), which tend to be red in the Donley et al. (2012)(see Eq.1and2 intheir paper),and presence of an AGN; v) using the depth of absorption considering the fluxes reported by U et al. (2012), we features (e.g., Imanishi & Dudley 2000; Risaliti et al. find that NGC6286 does not satisfy the conditions for 2006; Georgantopouloset al. 2011b), with large depths the presence of an AGN. The fact that the IR proxies pointing towards AGN obscured by dust; and/or vi) fail to identify the AGN emission in NGC6286 is due from deviations of the well known correlation between to the problematic identification of low-luminosity AGN the far-IR (FIR) and the radio luminosity (Helou et al. with IR spectra dominated by the host. For example, 1985, Condon et al. 1991, Condon 1992), using the in a low-luminosity AGN the silicate absorption feature radio-FIR flux ratio q (e.g., Yun et al. 2001). We find wouldbe dilutedbythe strongIRcontinuumofthe host that all these proxies (Table2) fail to detect the AGN galaxy. in NGC6286, with the exception of the NeV lines. Iwasawa et al. (2011) studied 44 LIRGs from the These lines can however also be produced by a young GOALSsamplewithChandra,andassessedthepresence starburst with a large population of Wolf-Rayet and O ofanAGNusingthehardnessratioHR≡(H−S)/(H+ stars (e.g., Abel & Satyapal 2008), so their detection S), where H and S are the background-correctedcounts does not always provide conclusive evidence of an AGN. in the 2–8 and 0.5–2keV ranges, respectively. Sources This is especially true for NGC6286, since the NeV with HR > −0.3 are reported as candidate AGN. This lines are weak [log(L /ergs−1) ∼ 38.8]. Moreover, value was chosen because ULIRGs which are known to [NeV] the detection of the [NeV] lines has been questioned host AGNs, such as Mrk231, Mrk273, and UGC5101, by Inami et al. (2013), who found [NeV]14.32µm tend to cluster just above this limit (Iwasawa et al. to be detected only in one of the two Spitzer ob- 2009). Considering the spatially-integrated X-ray flux servations available, while in both observations a NGC6286 has a hardness ratio HR = −0.85 ± 0.07, feature possibly consistent with [NeV]24.32µm were which would not allow to infer the presence of an AGN. However, as discussed by Iwasawaet al. (2011) NuSTAR unveils a low-luminosity heavily obscured AGN in the LIRG NGC 6286 9 this threshold could become less reliable for some CT Comparing this to the IR luminosity of the system AGN, since mostly reprocessed radiation is observed in [log(L /ergs−1) = 44.96] we find that the IR lumi- IR the hard X-ray band. Another criteria commonly used nosity of the AGN is between 0.1 and 0.6% of the total to identify AGN is the observed 2–10keV X-ray lumi- IR luminosity. This value is in disagreement with that nosity. Using log(L2−10/ergs−1) > 42 as a threshold obtainedbyVega et al.(2008)usingspectraldecomposi- (e.g., Szokoly et al. 2004, Kartaltepe et al. 2010), one tion, who found that the contribution of the AGN to would also miss identifying NGC6286 as a buried AGN the total IR luminosity is about one order of magni- [log(L2−10/ergs−1)=40.80]. tude larger. A 5% contribution to the total IR lumi- Spectral decomposition (e.g., Nardini et al. 2008, nositywouldimplythatlog(LAGN/ergs−1)=43.66and IR Alonso-Herrero et al. 2012) is another powerful method the intrinsic 2–10keV luminosity of the AGN would be to constrain the contribution of AGN to the multi- log(L2−10/ergs−1) = 43.12, also an order of magnitude wavelength SED. Vega et al. (2008) found that a pure larger than predicted by our X-ray spectral analysis. To starburst model fails to reproduce well the near-IR to have such a luminosity, the AGN should be obscured by radio SED of NGC6286, and a buried AGN account- log(N /cm−2) > 25, which is inconsistent with the re- H ing for 5% of the IR luminosity is required by the sults obtained here. An alternative explanation for this data. An useful diagnostic of the presence of a heav- discrepancy is that the AGN is intrinsically weak at X- ily obscured AGN is the ratio between the MIR and ray wavelengths, as recently found by NuSTAR for the the 2–10keV luminosities (e.g., Alexander et al. 2008; AGN in Mrk231 (Teng et al. 2014, see also Teng et al. Rovilos et al. 2014; Georgantopoulos et al. 2011a). It 2015). has been shown indeed that for AGN the absorption- Assuming a 2–10keVbolometriccorrectionofκ =20 x corrected 2–10keV and the 6 and 12µm luminosities (e.g., Vasudevan & Fabian 2007), the bolometric output are well correlated (e.g., Gandhi et al. 2009; Stern 2015; of the AGN would be 7 − 40×1042ergs−1. This im- Asmus et al. 2015), so that deviations from the corre- plies that the ratio between the IR luminosity and the lation might imply the presence of heavy obscuration. total output of the AGN is LBol /L ≃ 0.8 − 4.5%. AGN IR Vega et al. (2008) report that at 6µm about 58% of the The AGN bolometric output can also be inferred from flux is produced by the AGN. This would imply that the [NeV]14.32µmluminosity, followingthe relationob- the ratio between the IR and observed X-ray AGN lu- tained by Satyapal et al. (2007): minosity is very low: L2−10/L6µm ≃ 2.4×10−3. This value is consistent with undetected DOGs in the CDF- logLABoGlN =0.938logL[NeV]+6.317, (2) N (Georgakakis et al. 2010) and with other U/LIRGs and is log(LAGN/ergs−1) ∼ 42.7, consistent with the (Georgantopoulos et al. 2011a), which is related to the Bol estimate obtained using the X-ray luminosity. The 2– fact that in U/LIRGs the IR emission is enhanced by 10keV bolometric correction obtained using this value strong star formation, leading to very low values of is κ ≃ 3 − 17. The black hole mass of NGC6286 Lity2−o1b0/taLin6µemd.inUs§i4ng(Lth2−e1l0ar∼ge2st×2–110042keerVgsX−-1r)ayonluemwionuolsd- Chaasraxbmeeetne &estBimieartmedannto(2b0e10M)uBsHing∼the2.b7la×ck1h0o8lMem⊙asbsy- stillfindthatL2−10/L6µm ∼0.1,avaluelowerthanthat spheroid correlation (e.g., Magorrianet al. 1998). The expectedfromthe L2−10−L6µm correlation. This might Eddington ratio of the source would then be λ ≃ imply that the AGN contribution to the IR flux is sig- Edd (0.2−1.2)×10−3, consistent with a low accretion rate nificantly lowerthan thatreportedby Vega et al.(2008) AGN. (see §5.2 and Fig.7). The lack of a significant AGN contribution to the to- talIRfluxisalsoconfirmedconsideringthe[NeV]/[NeII] 5.2. AGN contribution to the IR luminosity ratioversustheEWofthe6.2µmPAHfeature[seeFig.1 The IR luminosity of NGC6286 is 8.8×1044ergs−1, and 2 of Petric et al. (2011)], which shows that the ra- which would imply that, depending on the X-ray spec- tio between LAGN and L is below 1% for this object. IR IR tral model used, we obtain a ratio L2−10/LIR ≃ 4 × This, together with the 2–10keV bolometric correction 10−4 − 2.3 × 10−3, significantly lower than the value obtainedusing [NeV]14.32µm,clearly disfavoursthe in- expected from pure AGN (e.g., Mullaney et al. 2011). trinsically X-ray weak AGN scenario. We can there- Considering the observed 2–10keV luminosity, the ra- foreconcludethattheenergeticsofNGC6286areclearly tio is log(Lobs /L ) ≃ −4.14, which is consistent dominated by the host galaxy, with the low-luminosity 2−10 IR with the average value found for the GOALS sample AGN providing only a minor contribution to the total [log(Lobs /L )=−4.40±0.63, Iwasawa et al. 2011]. flux. The contribution of the AGN to the IR flux of the 2−10 IR Using the relationship of Mullaney et al. (2011), it is system is shown in Fig.7. possible to convert the 2–10keV luminosity into the ex- pected IR luminosity emitted by the dust around the 5.3. Optical and radio emission AGN: NGC6286 has been classified as a low- ionization nuclear emission-line region (LINER) by logLIARG,N43 =(0.53±0.26)+(1.11±0.07)logL2−10,43. Veilleux & Osterbrock (1987) using a classification (1) schemebasedonthe diagramfirstproposedbyBaldwin, In the above equation LAIRG,N43 and L2−10,43 are the 8– Phillips&Terlevich(1981). While mostLINERsappear 1000µm and 2–10keV luminosities of the AGN in units to be driven by old stellar populations (e.g., Sarzi et al. of1043ergs−1. Consideringthe rangeof valuesobtained 2010) and by shocks in ULIRGs (e.g., Soto & Martin for the 2–10keV intrinsic luminosity, the IR luminos- 2010, 2012), in some cases they can be associated ity of the AGN is log(LAGN/ergs−1) = 41.91−42.75. to low-luminosity AGN (e.g., Ho 2008). Yuan et al. IR 10 Ricci et al. BGSsamplefor11<log(LIR/L⊙)<12(DAGN ≃0.35). ThedensemoleculargastracerHCNhasbeenfoundto beenhanced(relativetoHCO+andCO)insystemswith dominantAGN(e.g.,Imanishi et al.2007). Privon et al. (2015)haveshownthatsomepurestarburstandcompos- ite sources show similarly enhanced HCN emission. The originofthisenhancementisuncertain,butmightbedue to mid-infrared pumping associated with a compact ob- scured nucleus (CON; e.g., Aalto et al. 2015). However, the HCN/HCO+ ratio of NGC6286 is consistent with that of normal starbursts,rather than CONs. From this we can conclude that the starburst does not appear to be compact. Aradiocoreisrathercommoninlow-luminosityAGN, as shown by the work of Nagar et al. (2005), who found Figure 7. IntrinsicX-rayluminosityoftheAGNinthe2–10keV evidenceofradioemissionin≥50%ofthelow-luminosity bandversusthetotalIRluminosityofthesystem(inthe8–1000µm AGN of the Palomar Spectroscopic sample (see also Ho band). Bothluminositiesareinunits of1043ergs−1. Thecontin- 2008). The flux of NGC6286 at 1.4GHz is f = uousblacklinerepresentsthevaluesforwhichAGNandstarburst 1.4GHz contribute in the same amount to the IR flux, while the dashed 157.4±5.6mJy(Condon et al.1998),whichimplies that linesshowcontributionsoftheAGNof20,10and1%. Thevalues the radio loudness is logRX = log(f1.4GHz/f2−10) = oftheIRluminosityexpectedtobeduetotheAGNarecalculated −2.6 to −3.1, depending on the X-ray spectral model fromthe2–10keVluminosityfollowingEq.1. Thetwovaluesofthe assumed. These values were obtained taking into ac- 2–10keVluminosityofNGC6286representtheminimumandmax- imumvalueobtainedwiththedifferentmodelsdiscussedinSect4.2 count only the nuclear emission in the computation of (see also Table3). The figure shows that the AGN in NGC6286 the 2–10keV flux. Considering the threshold suggested contributes <1%ofthetotalIRluminosity. by La Franca et al. (2010) (see also Panessa et al. 2007, Terashima & Wilson 2003), logR = −4.3, NGC6286 X (2010) have recently used a new semi-empirical optical would be classified as radio-loud AGN. Murphy (2013) spectral classification to classify IR-selected galaxies report that the radio spectral index19 of NGC6286 is based on three diagrams: [OIII]/Hβ versus [NII]/Hα, α = −0.73±0.03, α = −0.89±0.03 and α = low mid high [SII]/Hα and [OI]/Hα line ratios. This is based on −1.02 ± 0.12 for ν < 5GHz, 1 < ν/GHz < 10 and the work of Kewley et al. (2006) to separate starburst ν > 10GHz, respectively. This would point towards a galaxies, starburst/AGN composite galaxies, Seyfert2s, significant contribution of synchrotron emission, possi- and LINERs. In the scheme of Kewley et al. (2006) bly from a jet. The two radio sources detected by EVN objects that were classified as LINERs according to and coincident with the 3–8keV core could be in fact Veilleux & Osterbrock (1987) would be either true associated to a jet and counter jet, consistent with the LINERs, Seyfert2s, composite HII-AGN galaxies, or radio-loudclassification of NGC6286. high metallicity star-forming galaxies. Yuan et al. (2010) found that true LINERs are rare in IR-selected 5.4. Heavily obscured low-luminosity AGN in U/LIRGs samples(<5%),andmostoftheobjectswouldbeeither classified as star-forming galaxies or starburst/AGN AsdiscussedaboveforthecaseofNGC6286,theiden- composites. Yuan et al. (2010) classified NGC6286 as tification of heavily obscured AGN in LIRGs can be a composite using [NII], a HII region using [SII] and a rather difficult if the AGN has a low-luminosity. The LINER using [OI]. Therefore they adopted a composite EW of PAH features would not be significantly affected classification for the source, which might imply the by the AGN if it is highly obscured, since the gas and presence of an AGN. Yuan et al. (2010) found that in dust wouldshield the PAH-emitting molecules, or if it is the IR luminosity bin LIR = 1011−1012L⊙ about 37% notveryluminous. Alow-luminosityAGNwouldalsobe of the objects in the IRAS BrightGalaxy Sample (BGS, difficult to findby studying the 2.5–5µmslope, since the Sanders et al. 1995, Veilleux et al. 1995) are classified IR emission would be dominated by the starburst, and as composites. the AGN emission can still be self-absorbed. Absorp- To characterise the relative AGN contribution to the tion features also would not be able to help if the AGN extreme ultraviolet (EUV) radiation field, Yuan et al. is not very luminous. A more reliable tracer is [NeV], (2010) use D , which is the normalised distance but while its detection might indicate the presence of AGN from the outer boundary of the star-forming sequence. an AGN, its non-detection does not exclude it. More- While this quantity does not provide information on the over, [NeV] could be created in young starbursts, and fraction of emission due to the AGN, it can be used for low-luminosity AGN it could be too faint to be de- to compare the amount of EUV radiation due to the tected(seeEq.2). Radiostudiescanalsogiveimportant AGN in different objects. For NGC6286 they found insights, but since not all AGN are very strong at these D = 0.5 using both the [OI]/H and the [NII]/H wavelengths, results are not always conclusive. Hard X- AGN α α diagram. Yuan et al. (2010) found a statistically signifi- ray studies are possibly the best way to unveil obscured cantincreaseofD withL ,consistentwiththeidea AGN in U/LIRGs, although they can also be limited by thatthefractionofAAGNGNincreaIRsesforincreasingvaluesof absorption for log(NH/cm−2)≫24. the 8-100µm luminosity (e.g., Veilleux et al. 1995). The value obtained for NGC6286 is marginally larger than 19Weconsiderherethefollowingdefinitionofthespectralindex: the averagevalue obtained by Yuan et al. (2010) for the Fν ∝να.