On the LINER nuclear obscuration, Compton-thickness and the existence of the dusty torus; Clues from Spitzer/IRS spectra

2.8 MB·English

by  O. Gonzalez-Martin

Astronomy&Astrophysicsmanuscriptno.OGM-MIDLINERs (cid:13)cESO2015
January19,2015
On the LINER nuclear obscuration, Compton-thickness and the
existence of the dusty torus
Clues from Spitzer/IRS spectra
O.González-Martín1,2,3(cid:63),J.Masegosa4,I.Márquez4,J.M.Rodríguez-Espinosa1,2,J.A.Acosta-Pulido1,2,
C.RamosAlmeida1,2,D.Dultzin5,L.Hernández-García4,D.Ruschel-Dutra1,2,6,andA.Alonso-Herrero7,8
1 InstitutodeAstrofísicadeCanarias(IAC),C/VíaLáctea,s/n,E-38205LaLaguna,Spain
2 DepartamentodeAstrofísica,UniversidaddeLaLaguna(ULL),E-38205LaLaguna,Spain
5 3 CentrodeRadioastronomíayAstrofísica(CRyA-UNAM),3-72(Xangari),8701,Morelia,Mexico
1 4 InstitutodeAstrofísicadeAndalucía,CSIC,GlorietadelaAstronomías/n18008,Granada,Spain
0 5 InstitutodeAstronomía,UniversidadNacionalAutónomadeMéxico,ApartadoPostal70-264,04510MéxicoDF,Mexico
2 6 DepartamentodeAstronomia,UniversidadeFederaldoRioGrandedoSul,9500BentoGoncalves,PortoAlegre,91501-970,Brazil
7 InstitutodeFísicadeCantabria,CSIC-UC,E-39005Santander,Spain
n
8 DepartmentofPhysicsandAstronomy,UniversityofTexasatSanAntonio,SanAntonio,TX78249,USA
a
J
Received31stOctober2014;accepted15thJanuary2015
5
1
ABSTRACT

]
A Context.Mostoftheopticallyclassifiedlowionisationnarrowemission-lineregions(LINERs)nucleihostanactivegalacticnuclei
(AGN).However,howtheyfitintotheunifiedmodel(UM)ofAGNisstillanopenquestion.
G
Aims.Theaimsofthisworkaretostudyatmid-infrared(mid-IR)(1)theCompton-thicknatureofLINERs(i.ehydrogencolumn
. densitiesofN >1.5×1024cm−2);and(2)thedisappearanceofthedustytorusinLINERspredictedfromtheoreticalarguments.
h H
p Methods.Wehavecompiledalltheavailablelowspectralresolutionmid-IRspectraofLINERsfromtheInfraRedSpectrograph(IRS)
onboardSpitzer.Thesamplecontains40LINERs.WehavecomplementedtheLINERsamplewithSpitzer/IRSspectraofPGQSOs,
-
o Type-1Seyferts(S1s),Type-2Seyferts(S2s),andStarburst(SBs)nuclei.WehavestudiedtheAGNversusthestarburstcontentinour
r sampleusingdifferentindicators:theequivalentwidth(EW)ofthepolycyclicaromatichydrocarbon(PAH)at6.2µm,thestrength
t
s ofthesilicatefeatureat9.7µm,andthesteepnessofthemid-IRspectra.WehaveclassifiedthespectraasSB-dominatedandAGN-
a dominated,accordingtothesediagnostics.Wehavecomparedtheaveragemid-IRspectraofthevariousclasses.Moreover,wehave
[ studiedthecorrelationbetweenthe12µmluminosity,νL (12µm),andthe2-10keVenergybandX-rayluminosity,L (2−10keV).
ν X
Results. In 25 out of the 40 LINERs (i.e., 62.5%) the mid-IR spectra are not SB-dominated, similar to the comparison S2 sam-
1
ple(67.7%).TheaveragespectraofbothSB-dominatedLINERsandS2sareverysimilartotheaveragespectrumoftheSBclass.
v The average spectrum of AGN-dominated LINERs is different from the average spectra of the other optical classes, showing a
6
rather flat spectrum at 6−28µm. We have found that the average spectrum of AGN-dominated LINERs with X-ray luminosi-
2
ties L (2−10keV)>1041erg/s is similar to the average mid-IR spectrum of AGN-dominated S2s. However, faint LINERs (i.e.
8 L (2−X10keV)<1041erg/s)showflatspectradifferentfromanyoftheotheropticalclasses.ThecorrelationbetweenνL (12µm)and
3 X ν
L (2−10keV)forAGNnicelyextendstowardlowluminositiesonlyifSB-dominatedLINERsareexcludedandthe2-10keVband
0 X
X-rayluminosityiscorrectedinCompton-thickLINERcandidates.
.
1 Conclusions.WehavefoundthatLINERsproposedasCompton-thickcandidatesatX-raywavelengthsmaybeconfirmedaccording
0 to the X-ray to mid-IR luminosity relation. We show evidence in favour of the dusty-torus disappearance when their bolometric
5 luminosityisbelowLbol(cid:39)1042erg/s.Wesuggestthatthedominantemissionatmid-IRoffaintLINERsmightbeacombinationofan
1 ellipticalgalaxyhost(characterisedbythelackofgas),astarburst,ajet,and/orADAFemission.Alternatively,themid-IRemission
: ofsomeofthesefaintLINERscouldbeacombinationofellipticalgalaxypluscarbon-richplanetarynebulae.Inordertoreconcilethe
v
Compton-thicknatureofalargefractionofLINERswiththelackofdusty-torussignatures,wesuggestthatthematerialproducing
i
X theCompton-thickX-rayobscurationisfreeofdust.
r Keywords. Galaxies:active-Galaxies:nuclei-infrared:galaxies
a
1. Introduction (UM)proposesthatbothtypesofAGNareessentiallythesame
objects viewed at different angles (Antonucci 1993; Urry &
Theemissioninactivegalacticnuclei(AGN)ispoweredbyac-
Padovani 1995). Anoptically thick dustytorus surrounding the
cretion onto a supermassive black hole (SMBH). AGN are tra-
central source would be responsible for blocking the region
ditionally divided into two main classes based on the presence
where these broad emission lines are produced (the broad line
(Type-1)ornot(Type-2)ofbroadpermittedlines(FWHM>2000
region,BLR)inType-2Seyferts.Thetorusmustnotbespheri-
kms−1)intheopticalspectrum.Theso-calledunificationmodel
callysymmetric,inordertoobscuretheBLR,whileallowingat
the same time the emission coming from region producing the
(cid:63) Juan de la Cierva Fellow and Tenure track at CRyA (e-mail:
[email protected])
Articlenumber,page1of20
A&Aproofs:manuscriptno.OGM-MIDLINERs
permitted narrow lines (known as narrow-line region, NLR) to does not suffer such a large extinction. Furthermore, the dust
reachusfromthesamelineofsight(LOS). that absorbs the shorter wavelength emission reradiates in the
Low-ionisationnuclearemission-lineregions(LINERs),first mid-IR and correspondingly produces a substantial fraction of
classified by Heckman (1980), are the dominant population of thebolometricfluxoftheobject.Dissectingthedetailedmid-IR
AGN in the local Universe (Ho et al. 1997). However, they re- spectra of AGN can reveal the properties of the dust in the nu-
main as one of the most captivating subsets of nuclear classes clearregion(e.g.Mendoza-Castrejónetal.2015).Subarcsecond
because their main physical mechanism is still unknown. The resolutionstudiesclaimatightcorrelationbetweenrestframelu-
nature of LINERs was initially sustained in their optical spec- minositiesνLν(12µm)andLX(2−10keV)forType-1andType-
trum,whichcanbereproducedwithavarietyofdifferentphys- 2 Seyferts (Horst et al. 2009; Gandhi et al. 2009; Asmus et al.
ical processes (e.g. photoionisation from hot stars, non-thermal 2011;Masegosaetal.2013;González-Martínetal.2013).Imag-
photoionisation,shocks,post-mainsequencestars,orAGN,Do- ingdataatmid-IRwavelengthshaveshownthatsomeCompton-
pita&Sutherland1995;Heckman1980;Ferland&Netzer1983; thick LINER candidates might fall off this relation, with L12µm
Veilleux&Osterbrock1987;Stasin´skaetal.2008;Singhetal. largerthanexpectedfortheirLX (Masonetal.2012;Masegosa
2013). In fact, many authors have suggested that LINERs are et al. 2013). The confirmation of such a trend for a large sam-
anheterogeneousclass(Satyapaletal.2005;Dudiketal.2005, ple might confirm the Compton-thick nature of a large portion
2009).Indeedtheemissionmechanismdominatingtheiroptical ofLINERs.
spectrum is still under debate. Singh et al. (2013) show that an From the theoretical point of view, Elitzur & Shlosman
AGNalonecannotexplaintheradialprofileofthesurfacebright- (2006)showedthatthetorusmightdisappearwhenthebolomet-
nessHαemissionline;acontributionofextendedemissionpost- ricluminositydecreasesbelowLbol ∼1042erg/sbecausetheac-
mainsequencestarsisneededatopticalfrequencies.Thereason cretionontotheSMBHcannotlongersustaintherequiredcloud
whythiscontributioncanbeseeninLINERsmightbetheintrin- outflow rate. Thus, the low bolometric luminosity of LINERs
sicweaknessoftheAGN,whichwouldoutshinethesesignatures makesthemkeytoprobethistheoreticalprediction.Maozetal.
inmorepowerfulAGN.Inlinewiththat,González-Martínetal. (2005) showed that the fraction of variable Type-1 and Type-2
(2014) showed that the host-galaxy contributes in a large frac- LINERsatUVisthesame.Thisfavoursthelackofadustytorus
tioninmostoftheLINERsevenatX-rays.Thus,theanalysisof obscuringthecentralAGNinType-2LINERssincethecentral
LINERs could be key to study the interplay between the AGN source is the responsible for such variability pattern. The mid-
andthehostgalaxy. IR emission shows clear signatures of the dusty torus. In fact,
Nowadays we know that around 75-90% of LINERs clumpy torus models (Nenkova et al. 2008) have succeeded in
show evidence of AGN using multi-wavelength information explainingthemid-IRemissionofSeyfertgalaxies(e.g.Ramos
(González-Martínetal.2006,2009A;Dudiketal.2009;Younes Almeidaetal.2009,2011;Alonso-Herreroetal.2011;Höniget
et al. 2011; Asmus et al. 2011; Mason et al. 2012; González- al.2010).Thus,themid-IRspectraofLINERscangiveimpor-
Martínetal.2014)1.ThisisconfirmedalsofromX-rayandUV tantcluesontheexistenceofthedustytorusforlowbolometric
variability studies (Maoz et al. 2005; Hernández-García et al. luminosities.However,othercontributorslikejetemission(e.g.
2013,2014).However,whatdoesmakeLINERsdifferentfrom NGC1052, Fernández-Ontiveros et al. 2012) or optically thin
other AGN? How do they fit into the UM of AGN? Some re- dust (e.g. NGC3998, Mason et al. 2013) can also contribute to
sults suggest that they might constitute a class of AGN with themid-IRemission.
a different accretion mode (e.g. Younes et al. 2011; Nemmen Thepurposeofthispaperistostudy(1)theCompton-thick
et al. 2014) while some other authors have argued that large nature of LINERs and (2) the plausible disappearance of the
obscuration is responsible for their differences (e.g. Dudik et torus.Wepresentthemid-IRSpitzer/IRSspectraof40LINERs.
al. 2009; González-Martín et al. 2009B). González-Martín et Wecomparethemwithmid-IRSpitzer/IRSspectraofSeyferts,
al. (2009A) found that the hydrogen column density, N , in PGQSOs,andStarbursts(seeSection2).Thedatareductionand
H
LINERs range from the galactic value up to N (cid:39)1024cm−2. measurementsaredescribedinSection3.Therelativelylowspa-
H
This is fully consistent with the N values reported for Seyfert tialresolutionofSpitzer/IRSspectramakesthesespectrabeing
H
galaxies (e.g. Panessa et al. 2006; Bianchi et al. 2012; Mar- contaminated from the host galaxy emission, which is partic-
inucci et al. 2012). However, using the ratio between the lu- ularly relevant for low-luminosity AGN (LLAGN) as LINERs.
minosity of the [OIII]λ5007Å emission line and the intrinsic Section4describesamethodwhichisabletoselectthosemid-IR
hard (2-10 keV) X-ray luminosity, L([OIII])/L (2−10keV), spectra with a negligible starburst contribution. Section 5 stud-
X
as a tracer of Compton-thickness (i.e. N >1.5×1024cm−2), iestheaveragespectrumofLINERsandcomparethemwiththat
H
González-Martínetal.(2009B)foundthatupto53%oftheLIN- ofSeyferts,PGQSOs,andStarbursts.Section6showstheanal-
ERsintheirsampleareCompton-thickcandidates.Thispercent- ysis of the correlation between LX(2−10keV) and L12µm for
age is two times higher than that reported for Type-2 Seyferts LINERs. In Section 7 we discuss the implications of the main
(Maiolino et al. 1998; Bassani et al. 1999; Panessa et al. 2006; results.TheconclusionsofthispaperaresummarisedinSection
Cappietal.2006).Dudiketal.(2009)studiedtheemissionlines 8.
in67high-resolutionSpitzer/IRSspectraofLINERsandfound
thatthecentralpowersourceinalargepercentageofLINERsis
2. Sample
highlyobscuredatopticalfrequencies,consistentwiththeX-ray
results. Our initial sample of LINERs comes from the catalog of
ObscuringdusthampersthestudiesoftheopticaltosoftX- LINERs observed at X-rays published by González-Martín
rayemissioncomingfrombothyounghot-starsandtheaccretion et al. (2009A). This guarantees that all the LINERs have
disk. However, emission at mid-infrared (mid-IR) wavelengths LX(2−10keV)measurements,whatiscrucialforourpurposes.
However,wemustbeawarethatthissampledoesnotconstitute
1 NotethatmostofthesestudiesselecttheirsourcesusingarchivalX- a complete sample. For the present analysis we have used two
raysobservationssotheymightbebiasedduetothecomplexselective databases to obtain the Spitzer data for an additional sample of
effectthisintroducesinthesample. LINERs.ThefirstoneistheCornellatlasofSpitzer/IRSspectra
Articlenumber,page2of20
O.González-Martínetal.:OntheLINERnuclearobscuration,Compton-thicknessandtheexistenceofthedustytorus
(CASSIS2).CASSISprovideslow-resolutionspectra(R∼60-127 Véron(2006)althoughitwasclassifiedasastarburstbyHo
over5.2µmto38µm)withtheIRSinstrumentsinthestaremode etal.(1997).
(Lebouteiller et al. 2011). The second database is the Spitzer
infrared nearby galaxy survey (SINGS Kennicutt et al. 2003). Notethatforalltheclasseswehaveonlyincludedspectraob-
SINGS is a Legacy programme of imaging and spectroscopic servedwithboththeshort-low(SL)andlong-low(LL)modules
data for 75 nearby galaxies. As part of the Legacy programme toguaranteethefullSpitzer/IRScoverage(atleast∼5−30µm).
all the one-dimensional nuclear spectra have been archived in Moreover,ultra-luminousinfraredgalaxies(ULIRGs)havebeen
the infrared science archive (IRSA3). This provides uniform 5- excludedfromtheanalysisbecausetheymighthaveacontrover-
30µmspectrainthespectralmappingmode.Mostgalaxiesinthe sialsourceofemissionatmid-IR(Imanishietal.2007;Alonso-
SINGSsamplehavealsobeenobservedwithChandraandtheir Herrero et al. 2013). All together these samples comprise 89
mainX-raypropertiesaredescribedbyGrieretal.(2011). sources(129nucleiincludingLINERs).
The CASSIS atlas contains 27 LINERs published by
González-Martínetal.(2009A).WehavealsoaddedNGC3079,
that was not analysed by González-Martín et al. (2009A) but 3. Dataprocessingandanalysis
it was included as a Compton-thick LINER by Goulding et al.
CASSIS and SINGS provide flux and wavelength calibrated
(2012). Grier et al. (2011) included 20 LINERs (8 included in
spectra.However,theobservationsusingdatafromboththeSL
González-Martínetal.2009A). and LL spectral modules suffer from mismatches due to tele-
The final sample of LINER in this paper contains 40 scopepointinginaccuraciesorduetodifferentspatialresolution
Spitzer/IRSspectra.Amongthemeighthavebeenopticallyclas-
oftheIRSorders.Thisisnotcorrectedinthefinalproductsgiven
sified as Type-1.9 LINERs (LINER1) and 32 as Type-2 LIN-
byCASSISandSINGS.Wethereforescaledeachspectratothe
ERs (LINER2) by Ho et al. (1997). Three of them are known immediateprior(inwavelengthrange)toovercomesucheffects.
Compton-thick,21areCompton-thin,and16wereclassifiedas
Thus, our flux level is scaled to the level of the shortest wave-
Compton-thickcandidatesbyGonzález-Martínetal.(2009B).
lengths, which is the order with the highest spatial resolution.
Dudik et al. (2009) reported mid-IR spectra of 67 LINERs
This guarantees that the flux level is scaled to the best spatial
(13 objects in common with our sample). However, they used
resolution that Spitzer can provide. Moreover, the spectra are
thehigh-resolutionmodesofSpitzer/IRSbecausetheirworkwas
shiftedtorest-frameaccordingtotheredshiftoftheobjects.
focused in the fine structure mid-IR emission lines. Thus, their
For each object we have measured the 12µm and 30µm lu-
resultsarenotdirectlycomparablewithours.Sturmetal.(2006) minosities using the Spitzer/IRS spectra. Errors have been esti-
reportedamid-IRstudyof33LINERs.Theyselectedtheirsam-
matedassuming15%flux-calibrationuncertainties,whichfully
pleonthebasisofIRluminositywhileoursampleisconformed
dominate other source of errors (e.g. González-Martín et al.
byobjectswithmeasuredX-rayluminosities.Asaconsequence,
2013;RamosAlmeidaetal.2011).Wehavealsomeasuredthe
onlyNGC4486iscommonwithoursample.
fluxesandequivalentwidth(EW)ofthepolycyclicaromatichy-
drocarbon (PAH) features at 6.2 and 11.3 µm. The EW of the
2.1. Comparisonsamples PAH features were measured by integrating the emission over
the continuum in a wavelength range of 5.9-6.5µm and 11.0-
To study the possible contribution of star-formation or AGN 11.6µmforthe6.2and11.3µmPAHemissionfeatures,respec-
emission,wehaveselectedstarbursts,SeyfertsandPGQSOsto tively. The continuum was estimated through a linear fit to the
be able to compare their mid-IR spectra to those of LINERs. 5.5-5.9µm (10.7-11.0µm) and 6.5-6.7µm (11.6-11.9µm) for the
Note that the sample is not complete in any sense but it allows 6.2µm (11.3µm) PAH feature (see e.g. González-Martín et al.
us to have a representative set of objects for each category to 2013).Wehavealsocomputedthestrengthofthesilicateemis-
compareLINERswiththem: sion/absorption feature at 9.7µm through the apparent depth at
9.7µm,τ (e.g.Shietal.2006;Levensonetal.2007):
9.7µm
– Seyferts.AlltheType-1andType-2sourcesincludedinShi
et al. (2006), in the Compton-thick sample described by
Goulding et al. (2012), and those included in the SINGS τ9.7µm =ln(Fcont,9.7µm/F9.7µm) (1)
sample. In total it contains 42 Seyferts. Among them 31
areType-2Seyferts(S2,including19Compton-thickand12 where F9.7µm and Fcont,9.7µm are the fluxes of the spectra around
Compton-thin)and11Type-1Seyferts(S1)4. 9.7µm and its expected continuum, respectively. Note that the
apparentdepthat9.7µmτ ispositiveforabsorptionsilicate
– Palomar GreenQSOs (PGQSOs). This sample includes all 9.7µm
featuresandnegativeforemissionfeatures.
thePGQSOsinthesampledefinedbyVeilleuxetal.(2009)
with Spitzer/IRS spectra in CASSIS and redshifts z<0.25. DuetothecomplexityofSpitzerspectra,wehaveusedPAH-
ThisPGQSOsampleincludes26sources. FIT6 toobtainτ9.7µm exceptwhenemissionsilicatefeaturesare
detected (see below). PAHFIT is an IDL tool for decompos-
– Starbursts. This sample is taken from Ranalli et al. (2003),
ing Spitzer/IRS spectra of PAH emission sources, with a spe-
Brandl et al. (2006), and Grier et al. (2011). The star-
cialemphasisonthecarefulrecoveryofambiguoussilicateab-
burst sample contains 21 sources. Note that among them
sorption,andweak,blendeddustemissionfeatures(Smithetal.
NGC3367 was classified as a Seyfert by Véron-Cetty &
2007).PAHFITisprimarilydesignedforitsusewiththefull5-
2 http://cassis.astro.cornell.edu/atlas/ 35µm Spitzer/IRS low-resolution spectra. However, PAHFIT is
3 http://irsa.ipac.caltech.edu not able to treat or recover silicate emission features expected
4 NotethattheS1samplecontainsobjectsclassifiedasType1,Type to occur in Type-1 AGN, giving τ9.7µm = 0. In these cases we
1.2,Type1.5,Type1.8,andType1.9Seyferts.TheS2sampleincludes have computed τ9.7µm by fitting the 9-14µm Spitzer/IRS spec-
onlypurelyType-2Seyferts. tratoaGaussianprofile.Thisisageneraltreatmenttoestimate
5 The redshift limit has been chosen to be able to obtain rest frame
30µmluminosities,requiredforouranalysis. 6 http://tir.astro.utoledo.edu/jdsmith/research/pahfit.php
Articlenumber,page3of20
A&Aproofs:manuscriptno.OGM-MIDLINERs
Fig. 1. Histograms of the EW of the PAH at 6.2 µm (left) and at 11.3 µm (right) for PGQSO (top-panel, orange-broad bars), S1 (top-panel,
red-narrowbars),SB(middle-panel,green-broadbars),S2(middle-panel,yellow-narrowbars),LINER1(bottom-panel,purple-narrowbars),and
LINER2(bottom-panel,pink-broadbars).Themedianvaluesand25th-75thpercentilerangeforeachclassofobjectareshownwithlargecrosses
(withthesamecolor-codethanthehistogram)withcontinuouslinesforPGQSO,SB,andLINER2andwithdashedlinesforS1,S2,andLINER1.
Theverticaldot-dashedlineshowsthelimitchosentodividethesampleintostrong-andweak-PAHs(seetext).
τ that has been proven to be a good approximation when tributionmightbeverystrongatmid-IRwavelengths,dominat-
9.7µm
comparedwithPAHFIT(seeGonzález-Martínetal.2013). ingtheentireemission(Masonetal.2012).
TogetherwiththeSpitzer/IRSspectraofthesamplewealso Overthelastdecadeseveraldiagnosticshavebeenproposed
compiled the 12µm luminosities obtained with ground based toquantifythecontributionofstarformationandAGNactivity
telescopes. These measurements have the advantage to better to the infrared luminosity. These diagnostics are based on the
isolatethenuclearsourcebecausetheycomefromimageswith mid-IR continuum slope, the EW of the PAH features, the ra-
∼0.3 arcsec spatial resolution (i.e. few tenths of parsecs for tioof[NeV](or[OIV])over[NeII],andtheEWofthePAHat
nearby galaxies). Most of these measurements come from the 6.2or11.3µmversusthe9.7µmopticaldepthτ9.7µm (Genzelet
catalog of sub-arcsecond mid-IR images of AGN reported by al. 1998; Lutz et al. 1998; Dale et al. 2006; Sturm et al. 2006;
Asmusetal.(2013).ItincludesSubaru/COMICS,VLT/VISIR, Spoonetal.2007;Baumetal.2010;Hernán-Caballero&Hatz-
Gemini/MICHELLE and Gemini/T-ReCS mid-IR data of 253 iminaoglou 2011). In this section we use several diagnostics to
objects. separatewhichSpitzer/IRSspectraarestronglycontaminatedby
We have also included four LINERs observed with non-AGN emission. Note, however, that the starburst contribu-
GTC/CanariCamasproprietarydata(programIDsGTC42-12B tion to the mid-IR spectra does not exclude the presence of an
and GTC35-13A). We have reduced them uniformly with the AGN.ThisisanattempttodeterminewhethertheAGNisdom-
RedCanpackage(González-Martínetal.2013).Wepresenthere inatingornotthemid-IRspectrum.Wehaveexcludedfromthe
their luminosities (marked with asterisks in Col. 5 in Table 1) analysis the diagnostics based on fine structure emission lines
whereas the full imaging analysis will be presented in a forth- becausetheyareblendedwithotheremissionlinesatthespectral
comingpaper.Alltogetherwehaveground-basedmeasurements resolutionofthesedataset.Formoredetailsinthesediagnostics
for61outofthe129sourcesincludedinthispaper. we refer the reader to Dudik et al. (2009), where they studied
Table 1 contains the X-ray luminosities, mid-IR measure- theseemissionlinesforalargesampleofLINERs.
ments for the Spitzer/IRS spectra, and the 12µm luminosities
fromSpitzer/IRSandground-basedtelescopeswhenavailable.
4.1. PAHfeatures
The star formation activity correlates with the PAH strength,
where starburst-dominated galaxies are then expected to show
4. AGNversusstarburstcontents
strong PAH features (e.g. Peeters et al. 2004). This well stab-
Spitzer hasbeenusedtostudythelargestsamplesofAGNever lishedcorrelationhasledtotheuseofPAHstrengthasatracer
analysedatmid-IR(e.g.Shietal.2006;Deoetal.2007).How- of star formation (e.g. Esquej et al. 2014). PAHs might be de-
ever, a disadvantage of these data is their relatively low spatial stroyedduetothepresenceofanAGN(Genzeletal.1998;Wu
resolution.ThismakesSpitzer spectratobeoftencontaminated etal.2009).ThisisparticularlyrelevantforthePAHfeatureat
by the host galaxy. It is expected to be particularly relevant for 6.2µm that is produced by grains with smaller sizes and, there-
LINERswheretheAGNisfaint.Inthiscasethenon-AGNcon- fore,theirdestructionneartheAGNismoreefficient(Diamond-
Articlenumber,page4of20
O.González-Martínetal.:OntheLINERnuclearobscuration,Compton-thicknessandtheexistenceofthedustytorus
Fig.2.Thesilicateapparentdepthat9.7µm,τ ,versustheEWofPAHat6.2µm,EW(PAH6.2µm),forPGQSOs,S1s,andSBs(left)andfor
9.7
LINERsandS2s,(right).LINER1s,LINER2s,S1s,S2s,PG-QSOs,andstarburstsareshownwithpurpleandpinkcircles,redandyellowsquares,
orangestars,andgreenup-sidedowntriangles,respectively.WhitecirclesandwhitestarsmarkknownCompton-thicksourcesandCompton-thick
candidates,respectively.Notethattheerrorbarsinτ arealwayswithinthesizeofthesymbol.IntherightpanelwealsoincludePGQSOs,S1s,
9.7
andSBswithwhitesymbolsforcomparisonpurposes.Green-dottedlinesindicatethediagonalbranchfoundbySpoonetal.(2007)forULIRGs
andStarbursts.Theshadowed(grey)areashowstherangeofvaluesforτ thatcouldbeexplainedwithClumpymodels(Nenkovaetal.2008).
9.7
Wealsoshowwithorangeandyellowlinestheexpectedrangeofvaluesforτ inface-onAGN(assumingi=0◦)andinedge-onAGN(assuming
9.7
i=90◦)usingthemodelsdescribedbyNenkovaetal.(2008)(seetext).Thered-shortverticallineshowsEW(PAH6.2µm)=0.228µmwhich
dividesintoweakandstrongPAHs.
Stanic & Rieke 2012). However, the 11.3µm might not be sup- histogramofEW(PAH6.2µm).Moreover,theS2classoverlaps
pressedbytheAGN(atdistancesascloseas10pc)butdiluted with the SB class for a larger number of objects in EW(PAH
when the AGN continuum becomes dominant (Alonso-Herrero 11.3µm) than for the histogram of EW(PAH 6.2µm) (15 and 6
et al. 2014; Ramos Almeida et al. 2014). Note that our aim is objects,respectively).
to select those spectra where the host galaxy contribution due ThePAHfeatureat11.3µmmightbestronglyattenuatedby
to star-formation is not dominating the mid-IR spectrum. The the silicate absorption feature (Brandl et al. 2006). González-
PAHstrengthisagoodtracerofstar-formationoccurringatfar Martínetal.(2013)estimatedthisattenuationtobeupto∼40%
distances from the AGN, where this destruction/dilution of the oftheintrinsicPAH featureat11.3µmfor τ =1.Thisper-
9.7µm
PAHfeaturesisnegligible.Supportingthis,thePAHat11.3µm centage can be higher for larger τ . The EW(PAH 6.2µm)
9.7µm
was negligible in 18 out of the 20 AGN reported by González- might be a better tracer of star-formation when the silicate at-
Martín et al. (2013) with high-spatial resolution spectra while tenuation is large because it is not embedded in the silicate ab-
their Spitzer spectra showed strong PAH features. Thus, these sorptionfeature.ThisisclearlyseeninFig.1,wheretheoverlap
diagnosticsarestillusefulinouranalysis,irrespectiveofthedi- of the S2 class (expected to be more attenuated than S1s under
lutionorsuppressionofthePAHfeaturesneartheAGN. theunifiedmodel)withtheSBsismuchhigherfortheEW(PAH
11.3µm)thanfortheEW(PAH6.2µm).Wethereforehavecho-
Fig.1showsthedistributionsoftheEW(PAH)perclassfor
senEW(PAH6.2µm)asabettertraceroftheSB-dominancein
the PAH features detected at 6.2µm (left) and 11.3µm (right).
oursample.
Notethatupperlimitsonthenondetectedlinesarenotincluded
We have defined a limit on EW(PAH 6.2µm) us-
in these histograms. The median EW(PAH 6.2µm) for the SB
ing the mean value and the standard deviation over this
class is significantly higher than that for the S1 and PGQSO
mean value for objects classified as SBs as follows:
classes. Only one S1 shows an EW(PAH 6.2µm) consistent <EW(PAH6.2µm)>−3×σ(EW(PAH6.2µm))=0.233µm(i.e.
with the SB class (NGC5033). Moreover, only one SB shows <log(EW(PAH6.2µm))>=−0.633).Thisensuresthat99.7%of
alimitontheEW(PAH6.2µm)consistentwithS1sorPGQSOs
SBsshowEW(PAH6.2µm)abovethislimit.Noteherethatthe
(NGC3184). The S2 class have objects with EW(PAH 6.2µm)
limitwouldbeEW(PAH6.2µm)=0.247ifthe99thpercentileis
overlappingwithvaluesfoundforS1s,PGQSOs,andSBs.LIN-
used.Thiswillgiveaslightlylessrestrictivelimit7.Above(be-
ERs(bothLINER1andLINER2),likeS2s,alsoshowEW(PAH
low) this value we classified the objects as strong-PAH (weak-
6.2µm) spreading a large range of values. The histogram of
EW(PAH11.3µm)issimilartothatofEW(PAH6.2µm).How- 7 Onlytwoobjectswillbeincludedintheweak-PAHclassifthelimit
ever, there is a larger overlapping between the distributions in is set to EW(PAH6.2µm)=0.247 compared to those obtained using
EW(PAH 11.3µm) for SBs with S1s and PGQSOs than in the EW(PAH6.2µm)>0.233,namelyNGC7130andNGC3367.
Articlenumber,page5of20
A&Aproofs:manuscriptno.OGM-MIDLINERs
PAH) objects. Six out of the 31 S2s are consistent with the
strong-PAH category; ten out of the 40 LINERs are classified
within the strong-PAH class. All of them are LINER2s except
NGC1097.
Spoon et al. (2007) presented a mid-IR diagnostic of the
AGN/ULIRGcontentbasedonτ versusEW(PAH6.2µm)(see
9.7
alsoHernán-Caballero&Hatziminaoglou2011).Theadvantage
ofthisdiagnosticisthatittakesintoaccounttheeffectsofstrong
obscuration of the nuclear source. They showed that galaxies
are systematically distributed along two different branches: (1)
a horizontal line with τ <1 of continuum AGN-dominated
9.7
to PAH-dominated spectra and (2) a diagonal line going from
deeply obscured (high τ and low EW(PAH 6.2µm) to PAH-
9.7
dominatedspectra(lowτ andhighEW(PAH6.2µm).Seyferts
9.7
and QSOs are found exclusively on the horizontal branch with
τ <1.ThelargemajorityofLIRGsandULIRGsinSpoonet
9.7
al. (2007) are located in the diagonal line. Starburst are placed
attheendofthetwobranches,withlargeEW(PAH6.2µm)and
τ <1. They argued that these two branches reflect a funda-
9.7
mentaldifferenceinthedustgeometryinthetwosetofsources.
The horizontal branch could have a clumpy structure while the
diagonalmightbesmooth.
Fig. 2 shows τ versus EW(PAH 6.2µm) for SBs,
9.7
S1s, and PGQSOs in the left panel and LINERs and
S2s in the right panel. PGQSOs and S1s (τ <1 and Fig. 3. Histograms of log(νLν(20µm)/νLν(30µm)) for PGQSO (top-
9.7
panel,orange-broadbars),S1(top-panel,red-narrowbars),SB(middle-
EW(PAH 6.2µm)<0.228µm) are clearly distinguished from
panel, green-broad bars), S2 (middle-panel, yellow-narrow bars),
SBs (EW(PAH 6.2µm)>0.233µm). This result is fully consis-
LINER1 (bottom-panel, purple-narrow bars), and LINER2 (bottom-
tent with that reported by Spoon et al. (2007). Our diagram
panel, pink-broad bars). The mean values and one standard deviation
showsveryfewnucleiwithdeepsilicatefeaturesandweakPAH over the mean for each class of objects are shown with large crosses
features. This was also found by Spoon et al. (2007) with only (with the same color-code than the histogram), continuous lines for
eightoverthe160objectsintheirsamplebelongingtothiscate- PGQSO,SB,andLINER2whiledashedlinesforS1,S2,andLINER1.
gory.WehavetestedtheuseoftheEW(PAH11.3µm)insteadof ThegreyareaoftheplotshowstherangeofvaluesexpectedforAGN
EW(PAH 6.2µm) in this diagram, finding a similar result. This according to the models given by Nenkova et al. (2008). The orange
was already reported by Hernán-Caballero & Hatziminaoglou andyellowverticallinesshowthesamerangesbutforinclinationan-
(2011)inalargesampleofSpitzer/IRSspectra. glesofi=0◦andi=90◦,assumingthatthesevaluesarerepresentative
offace-onandaedge-onAGN(seetext).
There is a maximum τ expected under the predictions of
9.7
the clumpy torus models for AGN. Larger values of τ can
9.7
beinterpretedassignificantcontaminationfromthehostgalaxy
(Alonso-Herreroetal.2011;González-Martínetal.2013).Inor-
dertoinvestigatethisissue,wehavecomputedτ usingasetof i=90◦(−0.86<τ <1.25).S1sarenaturallyexplainedwithin
9.7 9.7
modelswithinthelibrariesofCLUMPY8.Theseconsistonaset therangeofvaluesofτ expectedundertheclumpytorusmod-
9.7
of spectral energy distributions (SEDs) using the AGN clumpy els.S2stendtoshowlargerτ thanS1s.OnlytwoS2sareout
9.7
torusemissiondescribedbyNenkovaetal.(2008).Theparam- oftheexpectedrangewithclumpytorusmodels.OnlytwoSBs
eterrangeschosenarethosereportedbyGonzález-Martínetal. and two LINERs show τ >1.25. Following this diagram, we
9.7
(2013).WehavedownloadedtheSEDsforawidthofthetoroidal havedividedoursampleintothreecategories:
distributionσ=45◦,aratiobetweentheouterandtheinnerra-
diusofthetorusr /r =200,anexponentialslopeoftheradial – Deep-Silicate:Strengthofthesilicatefeatureabovethemax-
out int
distribution of clouds q=2, an optical extinction of the clouds imumexplainedbyclumpymodels(i.e.τ9.7 >1.25)regard-
rangingτ =5−150,andanumberofcloudsalongtheequator lessofEW(PAH6.2µm).
V
ofthetorusNo =2−20clouds.Thenumberofcloudsalongthe – Strong-PAH:τ9.7 <1.25andEW(PAH6.2µm)>0.233µm.
LOS, N, depends on the inclination angle, i, as N=N e−i2/σ2. – Weak-PAH:τ9.7 <1.25andEW(PAH6.2µm)<0.233µm.
o
WereferthereadertoGonzález-Martínetal.(2013)fordetails
This classification is included in Table 1. Starburst are lo-
ontheselectionoftheseparametersandtoNenkovaetal.(2008)
catedmostlyintheregionofstrong-PAH.Twoofthem,though,
forthedetailsonthemodelling.
populatetheareaofdeep-silicates.PGQSOsandS1sareplaced
We have computed τ9.7 for these SEDs using the same intheregionofweak-PAHs9.
methodology as for the Spitzer/IRS spectra reported here. Fig.
S2stendtoshowlargerτ thanS1s,asexpectedunderthe
9.7
2 shows as a grey area the minimum and maximum τ found
9.7 unifiedmodelofAGN.SimilarlytoS1sandPGQSOs,mostS2s
using these models (−1.13<τ <1.25). Thus, objects with
9.7 are in the region of weak-PAH. Only two of them are within
τ >1.25 are not expected under any clumpy torus model.
9.7 theareaofdeep-silicatesandfiveofthemarewithintheareaof
We use τ >1.25 to classify an object as deep-silicate. We
9.7 strong-PAHs.Thus,accordingtothisdiagram,onlysevenoutof
also show the range of τ expected for face-on AGN assum-
9.7 the31S2sshowsignsofhost-galaxycontaminationatmid-IR.
ingi=0◦(−0.96<τ <0.62)andforedge-onAGNassuming
9.7
9 TheonlyexceptionisNGC5033whichislocatedintheregionofthe
8 http://www.pa.uky.edu/clumpy/ diagramofstrong-PAHs.
Articlenumber,page6of20
O.González-Martínetal.:OntheLINERnuclearobscuration,Compton-thicknessandtheexistenceofthedustytorus
Fig.4.The12µmfluxobtainedwithground-basedtelescopes(nuclear)versus12µmfluxobtainedwiththeSpitzer/IRSspectra.Notethatboth
quantitiesareshowninlog-scales.Acrossisshowntoillustratetheerrorbarsinthesemeasurements(seeSection3).Theleftpanelshowsthe
resultsfortheentiresampleofobjectswithground-basedtelescopes,themiddlepanelshowstheobjectsclassifiedasSB-dominatedandtheright
panelshowsonlyobjectswithSpitzer/IRSspectraclassifiedasAGN-dominatedinthiswork(seetext).Thebluecontinuouslinerepresentsthe
one-to-onerelationandthegreendashedlinethebestlinearrelation.
LINERs mostly populate the area of weak-PAHs. All the Followingthesameideathanintheprevioussubsection,we
LINER1sbutNGC1097areinthisareaofthediagram.Among havecomputedlog(νL (20µm)/νL (30µm))oftheoreticalmod-
ν ν
LINER2s,twoareintheregionofdeep-silicatesandninearein elsobtainedwiththeClumpylibraries.Thevertical-thickorange
the strong-PAHs area. If LINERs in the weak-PAH area of the andyellowareas(Fig.3)showtherangeofsteepnessexpected
diagram are considered as AGN-dominated at mid-IR, then 30 from these models for face-on (assuming i = 0) and edge-on
outofthe41areAGN-dominatedatmid-IR. (assuming i = 90) torii, respectively (delimited also with verti-
calorangeandyellow,respectively).Theexpectedrangeofval-
ues found for the models of Type-1 AGN is almost identical to
4.2. Steepnessofthemid-IRspectra
therangeofvaluesfoundforPGQSOsandS1s.Moreover,the
steepnessofthespectrainthemodelofType-2AGNisexpected
The steepness of the mid-IR spectra characterises the relative
toincludethesamerangeofvaluesthanS1sbutextendedtoward
contribution of warm and cool dust to the mid-IR (Baum et
lower values. Thus, Type-2 AGN are expected to show steeper
al. 2010). Note that we refer here to the steepness at the spa-
tialresolutionsoftheSpitzer/IRSspectra(i.e.kpcscales)while spectra than Type-1 AGN. This is in agreement with our re-
sults,whereS2stendtoshowlowerlog(νL (20µm)/νL (30µm))
at smaller scales (below 100 pc scales obtained with ground- ν ν
than S1s. Seven out of the 20 SBs have a steepness fully
basedinstruments),themid-IRemissionisexpectedtobedom-
consistent with the model of Type-2 AGN. All the objects
inatedfromdustheatedbytheAGN(Hönigetal.2011;Ramos
below the minimum steepness predicted for Type-2 AGN
Almeida et al. 2011). It has been proven as a good indicator of
by the Clumpy libraries can be considered as SB-dominated
thestarburstcontentgivenitscorrelationwithEW(PAH11.3µm)
(log(νL (20µm)/νL (30µm))=−0.24). However, there are SBs
(e.g. Wu et al. 2009; Weedman et al. 2005; Brandl et al. 2006; ν ν
abovethatlimit.Thus,log(νL (20µm)/νL (30µm))<−0.24in-
LaMassa et al. 2012). This steepness has been defined in sev- ν ν
eralbandsbydifferentauthors;e.g.20-30µm(Baumetal.2010; dicatesthatthespectrumisSB-dominatedbutwecannotdiscard
thatspectrashowinglog(νL (20µm)/νL (30µm))>−0.24might
Weedmanetal.2005)or15-30µm(Brandletal.2006;Wuetal. ν ν
alsobeSB-dominated.
2009; Nardini et al. 2008). We present the 20 and 30µm lumi-
nosities to produce an estimate of the steepness of the mid-IR
We have classified as SB-dominated spectra those showing
spectrainoursample.
log(νL (20µm)/νL (30µm))<−0.24. Among the S2s, six ob-
ν ν
Fig. 3 shows the histogram of such steepness, expressed as
jects are therefore SB-dominated, three of them already classi-
log(νL (20µm)/νL (30µm)) (see also Table 1). PGQSOs and
ν ν fied as SB-dominated according to the strengths of the PAHs
S1s tend to show larger values of log(νL (20µm)/νL (30µm))
ν ν andsilicatefeatures.Combiningbothmethodstogether,tenout
than SBs. However, these two distributions overlap in the
ofthe31S2s(32%)areSB-dominated.
range −0.25<log(νL (20µm)/νL (30µm))<0. S2s and LIN-
ν ν
ERs show a wide range of log(νL (20µm)/νL (30µm)), over- Thirteen out of the 41 LINERs show
ν ν
lapping with S1s, PGQSOs, and SBs, although S1s and log(νL (20µm)/νL (30µm))<−0.24. Two are LINER1s
ν ν
PGQSOs distributions are skewed toward the larger values and the remaining 11 are LINER2s. Among them, eight were
of log(νL (20µm)/νL (30µm)). Therefore, this ratio itself is alreadyclassifiedasSB-dominatedusingthestrengthsofPAHs
ν ν
not as good tracer of AGN dominance as it is EW(PAH and silicate features. Interestingly, only two LINERs classified
6.2µm ) (see previous subsection). Although in theory it is a as SB-dominated with the diagram seen in Fig. 2 are not
good tracer of the contribution of warm and cool dust to the SB-dominatedusingthesteepnessofthespectra.Fifteenoutof
mid-IR, in practice, some SB-dominated spectra can have a the40LINERs(37.5%)showsignaturesofbeingSB-dominated
log(νL (20µm)/νL (30µm))ratioconsistentwiththoseofAGN. once the two methods presented in this section are considered
ν ν
Articlenumber,page7of20
A&Aproofs:manuscriptno.OGM-MIDLINERs
Fig.5.(Left):averagespectraforPGQSOs(orange),S1s(red),AGN-dominatedS2s(yellow),andAGN-dominatedLINERs(purple).(Right):
average spectra for SBs (green), SB-dominated S2s (yellow), and SB-dominated LINERs (purple). We also show one standard deviation as a
shadedregionusingthesamecolours.Eachspectrumisalsomarkedwithdifferentsymbolsat27µmforclarityoftheplot:PGQSOs(orangestar),
S1s(redsquare),SBs(greentriangle),S2s(yellowsquare),andLINERs(purplecircle).Theaveragespectraarescaledtothefluxat15µm.
together. The fraction of SB-dominated LINERs is similar to
thatofS2s. comparedtotheground-basedmeasurementsandbotharetrac-
inganuclearSB-dominatedspectrum.
If we select only AGN-dominated Spitzer spectra, the cor-
4.3. Goodnessofthemethodologytotracenuclear relation between these two quantities improves (see Fig. 4,
properties right panel) with a Pearson correlation coefficient of r=0.88
(P(null)=5.7×10−14). The linear fit to the AGN-dominated
We have considered a mid-IR spectrum as AGN-dominated sources (dashed line in Fig. 4, right panel) is very close to the
if it obeys these three criteria: EW(PAH 6.2µm)<0.233µm, one-to-onerelation(continuouslineinFig.4,rightpanel).The
τ9.7 <1.25, and log(νLν(20µm)/νLν(30µm))>−0.24. To study onlytwooutliersareNGC4594andNGC5866.Thus,whenthe
the goodness of this method to select AGN-dominated AGN-dominatedspectraareselectedthenuclear12µmfluxob-
Spitzer/IRS spectra, we have plotted in Fig. 4 the 12µm flux tainedwithground-baseddataisveryclosetothevalueobtained
fortheSpitzer/IRSspectraversusthesamequantityforground- by the Spitzer/IRS spectra. This reinforces our methodology as
based measurements for the 63 objects for which these mea- agoodtooltoisolateAGN-dominatedmid-IRSpitzer/IRSspec-
surements are available (see Section 3). Ground-based and tra.
Spitzer/IRS 12µm fluxes show a linear relation although the
dispersion is high (Pearson correlation coefficient of r=0.64,
P(null)=1.6×10−7). Moreover, the slope of the best linear fit 5. Averagespectra
(dashedlineintheleftpanelofFig.4)isflatterthantheone-to-
Fig. 5 shows the average spectrum for each class of objects.
onerelation(continuousline).
These average spectra have been computed after normalising
Most of the Spitzer/IRS spectra with larger 12µm flux than them to the flux at 15µm. The shaded regions show the stan-
those from ground-based measurements are SB-dominated ac- dard deviation over the average spectrum. We have computed
cording to the EW(PAH 6.2µm) method (see middle panel of the mean value for S2s and LINERs according to our mid-IR
Fig. 4). However, not all the SB-dominated Spitzer/IRS ob- classification(seeprevioussection)asAGN-dominatedandSB-
jects show a 12µm flux excess in the Spitzer/IRS spectra com- dominated(leftandrightpanels,respectively).
pared to the ground-based measurements. We rule out the ex- The average spectra of S1s and PGQSOs show similar
planation of a distance effect in which more distant objects shapes, showing the silicate feature in emission and similar
might include more SB-contribution in the nuclear spectra, be- steepness of the spectra (see left panel in Fig. 5). The relative
causeallourobjectsarenearbyandnoparticulartrendisfound differencesbetweenthesetwoclassesareanenhancementofthe
comparingSB-andAGN-dominatedsources.Alternatively,this strength of the silicate feature in emission for PGQSOs com-
result might have two explanations: (1) our method to select pared to S1s and a slightly steeper spectrum for S1s compared
SB-dominated spectra is too restrictive, and could include SB- to PGQSOs. The average SB spectrum is very different to that
dominated spectra that are actually AGN-dominated at 12µm, of S1s or PGQSOs (see right panel of Fig. 5). The main dif-
and (2) the Spitzer/IRS spectra do not contain extra emission ferences are strong PAH features, a steep spectrum, and deep
Articlenumber,page8of20
O.González-Martínetal.:OntheLINERnuclearobscuration,Compton-thicknessandtheexistenceofthedustytorus
different. This support our method as a good diagnostic of SB-
dominatedmid-IRspectra.ThiswasalsosuggestedbyAlonso-
Herreroetal.(2014),findingthattheSpitzerspectraofS1sand
S2saresimilaronlyifspectrawithdeepabsorptionsilicatefea-
tures are excluded from the analysis. The [Ne V] at 14.3µm
and[OIV]at25.9µmemissionlinesareclearlyseenintheav-
erage spectra of SB-dominated S2s and LINERs. This was al-
readyshownbyDudiketal.(2009),findingtheselinesinalarge
fractionofLINERs.Thus,theaverageSB-dominatedspectraof
LINERsandS2smightstillshowmid-IRsignaturesofAGNna-
ture,althoughtheoverallmid-IRspectraisnotdominatedbythe
AGN.
TheaverageS2AGN-dominatedspectrum(yellowspectrum
intheleftpanelofFig.5)doesnotmimicS1sorPGQSOs.This
averagespectrumissteeperthanthoseofS1sorPGQSOs.Italso
showsthesilicatefeaturesinabsorptionwhileS1sandPGQSOs
showanaveragespectrawithsilicatefeaturesinemission.This
is expected since the silicate feature at 9.7µm and 18µm are
predicted to be in emission for Type-1 AGN and in absorption
forType-2AGN(Nenkovaetal.2008).Thesepredictionshave
already been confirmed by observations (e.g. Shi et al. 2006).
Onthesimilarities,theaverage(AGN-dominated)S2spectrum
showsthepresenceof[NeV]at14.3and24.3µmand[OIV]at
25.9µmemissionlines.
TheaverageAGN-dominatedLINERspectrum(purplespec-
trumintheleftpanelofFig.5)canbeclearlydistinguishedfrom
PGQSOs, S1s, S2s, and SBs. The main characteristic of this
average spectrum is the rather flat continuum all over the 6-28
µm wavelength range. Moreover, it shows strong PAH features
at 11.3µm and 17µm. The [O IV] at 25.9µm emission line is
prominent as in S1s, PGQSOs, and S2s. However, the [Ne V]
at 14.3µm emission line is clearly undetected as is the [Ne V]
at 24.3µm. Below 20µm this spectrum resembles that of SBs.
However, it can be clearly distinguished from SBs because the
averagespectrumofLINERsdonotshowasteepspectrumand
it lacks of the silicate absorption features seen in SBs. More-
over, LINERs also show the [O IV] at 25.9µm that the SBs do
notshow.
5.1. Sub-classesofLINERs
Asexplainedintheintroduction,theLINERsareaheterogenous
Fig. 6. The average spectra of AGN-dominated LINERs according to family of objects. In order to study the subclasses of LINERs,
differentsubclassifications.Fromtoptobottom:(a)LINER1s(lightma- Fig.6showstheaveragespectraforseveralsubclassificationsof
genta)andLINER2s(darkmagenta);(b)brightLINERs(darkgrey)and LINERs. Note that these average spectra have been computed
faintLINERs(lightgrey);(c)Compton-thin(lightgreen)andCompton- includingonlyAGN-dominatedspectra.
thickcandidates(darkgreen);(d)objectsclassifiedatX-raysasAGN
The average spectrum of objects optically classified as
(dark blue) and non-AGN candidates (cyan). The average spectra for
LINER1s(panel(a),lightmagentaspectruminFig.6)issteeper
S1s(red)andAGN-dominatedS2s(yellow)arealsoshownforcompar-
than that of LINER2s (panel (a), dark magenta spectrum). In
isonpurposes.
fact,theaveragespectrumofType-1LINERsisconsistentwith
that of AGN-dominated S2s (yellow spectrum). However, the
dispersion in the average spectrum of LINER1s is quite large
because,amongthefiveAGN-dominatedLINERs1,NGC4450
silicateabsorptionfeatures.Moreover,theclassicallinesassoci-
shows a flat spectrum which is not consistent with the average
atedwithAGNemissionsuchas[NeV]at14.3µmand24.3µm
trendforthisclass.
or[OIV]at25.9µm,areclearlydetectedintheaveragespectra
We have also classified the AGN dominated LINERs into
ofPGQSOsandS1sbutareabsentintheaverageSBspectrum.
twoclassesattendingtotheirL (2−10keV):brightLINERsfor
NoteherethattheassociationoftheselineswithAGNemission X
objectswithL (2−10keV)>1041ergs−1andfaintLINERsfor
have been questioned, finding them in some SB galaxies (see X
thosewithX-rayluminositiesbelowthatlimit.Thisisthelimit
Pereira-Santaellaetal.2010).
wherethetorusisexpectedtodisappearatabolometricluminos-
The average SB-dominated spectrum for S2s and LINERs ity of L ∼1042ergs−1(Elitzur & Shlosman 2006)10. Among
bol
(rightpanelofFig.5)areverysimilartothatoftheSBs.Theav-
erage AGN-dominated (left panel in Fig. 5) and SB-dominated 10 This assumes a conversion between the X-ray luminosity and the
(right panel in Fig. 5) spectra for S2s and LINERs are clearly bolometricluminosityofL (cid:39)10×L (2−10keV)(Ho2008).This
bol X
Articlenumber,page9of20
A&Aproofs:manuscriptno.OGM-MIDLINERs
Fig.7.The2-10keVluminosityversusthe12µmluminosity,bothinlogarithmicscaleforPGQSOs,S1sandS2s(a),SBs(b),AGN-dominated
LINERs(c)andSB-dominatedLINERs(d).Thecontinuous-redanddashed-greenlinesshowthebestfitcorrelationsforAGNandSBs,respec-
tively,reportedbyAsmusetal.(2011).Thetypicalerrorforthesemeasurementsisshownasacrossinthetop-leftcornerofpanel(c).Errorsfor
theX-rayluminosityareestimatedas10%ofitsvalue.Thedot-dashedbluelineandlong-dashedlightgreenlineshowthelinearfitforPGQSOs,
S1s,andS2sandforSBs,respectively.Greyarrowsmarkobjectswithreportedupper-limitsontheX-rayluminosity.
theAGN-dominatedLINERs,7areclassifiedasbrightLINERs dispersion of the average spectrum of bright LINERs is much
and 18 are classified as faint LINERs. The resulting average lower than that of LINER1s. We have investigated if a differ-
spectra of bright and faint LINERs (dark and light grey spec- entmorphologyofthehostgalaxyforbrightandfaintLINERs
tra in panel (b) of Fig. 6) are quite similar to that of LINER1s couldproducethedifferencesintheiraveragemid-infraredspec-
and LINER2s. Thus, bright LINERs show a steeper spectrum tra.Interestingly,NGC4450istheonlyAGN-dominatedType-1
compared to faint LINERs, compatible with S1s. Note that the LINER hosted in a late-type galaxy. Only four (out of the 18)
faint LINERs and 1 (out of the 7) bright LINERs are hosted in
conversion factor is the most conservative value we have found (Ho latetypegalaxies(i.e.t>1).Thislackoflatetypegalaxieshost-
2009).Notethatanyhighervaluecouldincludemoreobjectsasbright ing LINERs is expected since they are usually hosted in early
LINERs(seealsothediscussion).
Articlenumber,page10of20