The nuclear dust lane of Circinus: collimation without a torus

1.9 MB·

by  M. Mezcua

Mon.Not.R.Astron.Soc.000,1–5(2013) Printed13January2016 (MNLaTEXstylefilev2.2)
The nuclear dust lane of Circinus: collimation without a torus
M. Mezcua1,2,3(cid:63), M. A. Prieto3,4, J.A. Ferna´ndez-Ontiveros5, K. R. W. Tristram6
1De´partementdePhysique,Universite´deMontre´al,C.P.6128,Succ.Centre-Ville,Montreal,QuebecH3C3J7,Canada
2Harvard-SmithsonianCenterforAstrophysics(CfA),60GardenStreet,Cambridge,Massachusetts02138,USA
6
3InstitutodeAstrof´ısicadeCanarias(IAC),E–38200LaLaguna,Tenerife,Spain
1
4UniversidaddeLaLaguna,Dept.Astrof´ısica,E–38206LaLaguna,Tenerife,Spain
0
5IstitutodiAstrofisicaePlanetologiaSpaziali(INAF-IAPS),ViaFossodelCavaliere100,00133Roma,Italy
2
6EuropeanSouthernObservatory,AlonsodeCo´rdova3107,Vitacura,Casilla19001,SantiagodeChile,Chile

n
a
J

1
1

ABSTRACT
]
A In some AGN, nuclear dust lanes connected to kpc-scale dust structures provide all the
extinction required to obscure the nucleus, challenging the role of the dusty torus proposed
G
by the Unified Model. In this letter we show the pc-scale dust and ionized gas maps of
.
h Circinusconstructedusingsub-arcsec-accuracyregistrationofinfraredVLTAOimageswith
p opticalHubbleSpaceTelescopeimages.Wefindthatthecollimationoftheionizedgasdoes
- not require a torus but is caused by the distribution of dust lanes of the host galaxy on
o
∼10 pc scales. This finding questions the presumed torus morphology and its role at parsec
r
t scales, as one of its main attributes is to collimate the nuclear radiation, and is in line with
s
interferometricobservationswhichshowthatmostofthepc-scaledustisinthepolardirection.
a
[ We estimate that the nuclear dust lane in Circinus provides 1/3 of the extinction required

to obscure the nucleus. This constitutes a conservative lower limit to the obscuration at the
1
centralparsecs,wherethedustfilamentsmightgetopticallythickeriftheyarethechannels
v
thattransportmaterialfrom∼100pcscalestothecentre.
3
5 Keywords: techniques:highangular resolution–astrometry– galaxies:nuclei–galaxies:
6
Seyfert–infrared:galaxies.
2
0
.
1
0
6 1 INTRODUCTION theobscuration/reflectionscenarioofAntonucci&Miller(1985),
1 the ionization cones result from shadowing of the nucleus by an
Supermassiveblackholesareubiquitousatthecentreofgalaxies
: optically-thick pc-scale torus. Many type 2 nuclei are Compton
v (McConnell&Ma2013),butonlyafractionofthemareactively
thick in the X-rays, a result often interpreted as further evidence
i accretingintheformofanactivegalacticnucleus(AGN).Those
X forapc-scaletorus(e.g.Risalitietal.1999).However,notalways
supermassive black holes powering an AGN are thought to be
do type 2 AGN present large X-ray column densities (N >
r surrounded on scales of a few parsec by a toroidal distribution H
a 1022 cm−2;e.g.Panessaetal.2009;Bianchietal.2012;butsee
ofwarmmoleculargasanddust(hencecalled“dustytorus”)that
Burtscher et al. 2015) and high N can be caused by the host
constitutes the reservoir of material that fuels the central engine. H
galaxyseenedge-on(e.g.Ho¨nigetal.2014).
According to the Unified Model of AGN (Krolik & Begelman
1988; Antonucci 1993; Urry & Padovani 1995) the dusty torus The torus is expected to be of parsec scales with a clumpy
is also responsible for the AGN dichotomy into type 1 (face-on morphology and be made of cold molecular material, a large
view, unobscured nuclei with broad and narrow optical emission fraction of it being dust particles that due to their proximity
lines)andtype2(edge-onview,obscurednucleiwithonlynarrow to the nucleus will absorb the UV emission from the accretion
optical emission lines) and for collimating the ionizing radiation. discandre-radiateitintheinfrared(IR;Krolik1998).CurrentIR
Evidence for this scenario relies on two compelling results: (i) interferometricobservationsnowfeasibleforthenearestAGNhave
spectro-polarimetric observations which reveal the presence of thecapabilitytoresolvethisstructure(e.g.Burtscheretal.2013).
broademissionlinesinsometype2AGN(e.g.Antonucci&Miller Instead of the expected geometrically-thick torus, interferometric
1985; Kay 1994) and (ii) the observed anisotropy of the nuclear observations reveal the presence of a mid-IR pc-scale polar
ionizingradiationfieldgivingrisetoconical-shapemorphologies elongationandadisc-likecomponentperpendiculartothesystem
of the narrow-line regions with the cone apex at the nucleus axis (i.e. in NGC1068, Jaffe et al. 2004, Lo´pez-Gonzaga et al.
(e.g. Tadhunter & Tsvetanov 1989; Prieto & Freudling 1993). In 2014); NGC424, Ho¨nig et al. 2012; NGC3783, Ho¨nig et al.
2013 and Circinus, Tristram et al. 2014). Extended mid-IR
emission in the polar direction has been also observed on
(cid:63) Email:[email protected] scales > 100 pc in the Circinus galaxy and a few other objects
(cid:13)c 2013RAS
2 Mezcuaetal.
(e.g.Packhametal.2005;Ho¨nigetal.2010;Reunanenetal.2010). subtracting the adjacent continuum following standard pure-line
Although a radiatively-driven dusty wind could explain such extractionprocedures.
polar-elongated dust structures (Ho¨nig et al. 2012, 2013), these
findings pose a further challenge to the role of the torus in the
frameworkoftheUnifiedModelandhowthisdisc-likecomponent 2.2 Radioobservations:thepc-scaleradiojet
plus extended polar dust is able to account for the obscuration
We searched for sub-arcsec resolution archival radio data of
-viewingangledependent-andthemorphologyoftheionizedgas.
Circinuswiththeaimofstudyingthemorphologyandorientation
In this letter we provide direct observational evidence that
of its radio jet at pc scales. We re-reduced archival 22 GHz data
a dusty torus is not required to explain the collimation of the
obtained in March 2005 with the Australia Telescope Compact
conical-like ionized gas structure seen in Circinus. Circinus is
Array (ATCA) using the MIRIAD package following standard
an almost edge-on (∼65 deg) SA(s)b galaxy, considered to be a
procedures. The data were then imaged with robust weighting,
prototype Seyfert 2 AGN with a narrow emission line spectrum
resultingina0.37arcsec×0.24arcsecbeamorientedataposition
(Oliva et al. 1994), polarized broad emission lines (Oliva et al.
angle(P.A.)of-8deg.ThecoreoftheradiojetofCircinus(central
1998)andaone-sidedionizationconethatextendsuptokpcscales
component in Fig. 2, bottom right) has a flux of 364.5 ± 4.9
inHαand[OIII](e.g.Veilleux&Bland-Hawthorn1997),∼60pc
mJy. The source is slightly resolved, with a size of 0.74 arcsec
inX-rays(Smith&Wilson2001)and∼30pcinthecoronalline
×0.52arcsec(measuredfromthe5σ contour)withanextension
[SiVII]2.48µm(Prietoetal.2005).ThenucleusisCompton-thick
toward the North-West along a P.A. of about -80 deg. This P.A.
(N ∼ 4 × 1024 cm−2; Matt et al. 1999) and only visible at
H is consistent with the direction of the polar emission detected by
wavelengthsλ > 1.6µm(Prietoetal.2004).Atadistanceof∼4
mid-IRinterferometry(P.A.=-76degNorthtoEast)androughly
Mpc(Freemanetal.1977),1arcseccorrespondsto∼19pc.
perpendiculartothenucleardustemitter(P.A.=46deg;Tristram
etal.2014).Sixmoreblobsaredetectedabovea5σlevelandare
most likely part of the large-scale radio lobes detected at lower
2 DATAANDANALYSIS frequencies(e.g.Elmouttieetal.1998).
2.1 IRadaptiveopticsandHSTopticaldata
Circinus was observed with the adaptive optics (AO) assisted 3 RESULTSANDDISCUSSION
near-IR instrument Naos-Conica (NaCo; Rousset et al. 2003) at
the ESO Very Large Telescope (VLT) as part of a larger sample 3.1 Thelocationofthenucleus
ofnearandbrightAGN.Thesampleanddatareductionhavebeen
The position of the nucleus can be identified as an outstanding
described elsewhere (Prieto et al. 2010; Reunanen et al. 2010;
point-like source in the NaCo 2µm Ks-band image, where it
Prieto et al. 2014; Mezcua et al. 2015). For the purpose of this
reaches its maximum contrast. This is particularly important for
Letter, we use the 2µm broad-band Ks filter (λ = 2.180µm,
c those galaxies hosting a type 2 AGN, in which the nucleus is
∆λ=0.350µm)toobtainthenear-IRcontinuumimage,asatthis
obscured either by a dust filament crossing the nuclear region
bandamaximumcontrastbetweentheAGNandthehostgalaxyis
(Prietoetal.2010,2014)orapc-scaletorus.Asaconsequence,an
observed(e.g.Prietoetal.2010,2014).TheNaCobeamresolution
offsetbetweentheKs-bandandopticalpeaksofemissionofseveral
intheKs-bandisof0.16±0.02arcsec.
tensofpcisobserved.SuchIR–opticalshiftwasfoundforsome
We use the archival Hubble Space Telescope (HST) broad
Sy2 galaxies (e.g. NGC1068, NGC1386 and NGC7582; Prieto
I-bandimagefromtheF814WfilteroftheWFPC21toconstructan
et al. 2014), including Circinus (Prietoetal.2004; Mezcua et al.
extinctionmapKs/I ordustmap.Whenthereisdustatalocation,
2015).Hereweunambiguouslyidentifythepositionofthenucleus
I is suppressed there whereas Ks is enhanced relative to other
of Circinus, revealed as a bright point-like source in the 2µm
locations where there is less or no dust. These enhanced regions
K-bandimages,withaprecisionof30mas.Thisnear-IRnucleus
show brighter in our Ks/I figure, indicating higher extinction
isfoundtobeshiftedby160mas±30mastothepeakofI-band
AV, and take the shape of filaments and lanes, suggesting emission,inagreementwiththe0(cid:48).(cid:48)15offsetderivedbyPrietoetal.
these are dust regions. The HST/F814W image corresponds to
(2004)andindicatingthatthenucleusisclearlyobscuredatoptical
the reddest HST optical image available and provides a large
wavelengths(seeFig.1,middleandrightpanels).
number of common reference sources with the NaCo/Ks-band
imagetobeusedfortheimagealignment.Theregistrationofthe
NaCo/Ks and HST/F814W images is reported in Mezcua et al. 3.2 Collimationbydustlanes
(2015) and was performed using the centroid position of three
The nucleus (or 2µm peak of emission) is located behind a dust
reference sources (the nucleus was not used; see Fig. 1, top left
filament or lane that extends up to ∼1.5 arcsec (∼30 pc; see
panel) to derive the relative offsets between the images. The
Fig. 1 right panel) forming a horned shape. This horn-shaped
alignmentuncertaintyisof30mas.Usingthesameprocedure,we
registertheWFPC2/F502N,WFPC2/F656NandNICMOS/F187N filament appears to delimit the sharp edges of the [OIII] cone
of emission as well as the main structure of the cone of Hα
filter images and their corresponding adjacent continuum to
the NaCo/Ks and HST/F814W images. Five common sources emission(Fig.2).AdditionalclumpyHαemissionisdetectedata
5σleveloutsidetheconedefinedbythehorn-shapedfilament.This
are used in this image registration. The alignment uncertainties
are of 10 mas, 40 mas and 20 mas for the WFPC2/F502N, clumpy Hα emission most likely traces low-level star formation
WFPC2/F656N and NICMOS/F187N alignments, respectively. relatively unobscured by foreground dust, as supported by its
The [OIII], Hα+[NII] and Paα images are then constructed by spatialcoincidencewithclumpywarmmolecularH21-0S(1)2.12
µmlinegasemissionNortheastofthenucleus(Fig.2,topmiddle;
Mezcuaetal.2015).Thesharpdefinitionoftheedgesofthe[OIII]
1 Wide-FieldPlanetaryCamera2. cone (Fig. 2, top left) being strictly traced at the vortex region
(cid:13)c 2013RAS,MNRAS000,1–5
Circinusdustlane 3
10 2 2
F814W + Ks F814W + Ks Ks/F814W + Ks
5 1 1
c) c) c)
e e e
cs cs cs
ar ar ar
ec. ( 0 ec. ( 0 ec. ( 0
D D D
e e e
v v v
ati ati ati
el el el
R R R
5 1 1
N
2" 0.5"
38 pc 10 pc E
10 2 2
10 5 0 5 10 2 1 0 1 2 2 1 0 1 2
Relative R.A. (arcsec) Relative R.A. (arcsec) Relative R.A. (arcsec)
Figure1.Left:HST/F814WimagewithNaCo/Ks-bandcontinuumcontoursinwhite.TheFoVis20arcsec×20arcsec.Bluecirclesmarkthepositionofthe
point-likesourcesusedforimagealignment.Theinner4arcsec×4arcsecregion(bluesquare)isshownindetailinthenextpanels.Middle:HST/F814W
imagewithNaCo/Ks-bandcontinuumcontoursinwhite.Right:Ks/F814WfluxratiowithNaCo/Ks-bandcontinuumcontoursinwhite.Thebrighterregions
intheKs/F814Wimagedenotehigherextinctionandhencedustabsorption.Thepositionofthenucleusanditserrorismarkedwithacrossinthemiddleand
rightpanels.TheangularresolutioninallpanelsisthatoftheNaCoKs-band(0.16arcsec).
2 2 2
ec) 1 ec) 1 ec) 1
s s s
c c c
ar ar ar
c. ( c. ( c. (
e 0 e 0 e 0
D D D
e e e
v v v
ati ati ati
el 1 el 1 el 1
R R R
0.5"
Ks/F814W + [OIII] 10 pc Ks/F814W + Hα +H2 Ks/F814W + [Si VII]
2 2 2
2 1 0 1 2 2 1 0 1 2 2 1 0 1 2
Relative R.A. (arcsec) Relative R.A. (arcsec) Relative R.A. (arcsec)
2 4
ATCA 22 GHz
ec) 1 ec) 2
s s
c c
ar ar
c. ( c. (
e 0 e 0
D D
e e
v v
ati ati
el 1 el 2
R R
Ks/F814W + Paα 0.5-7 keV
2 4
2 1 0 1 2 4 2 0 2 4
Relative R.A. (arcsec) Relative R.A. (arcsec)
Figure2.Ks/F814Wfluxratiowithemissionlinecontoursof[OIII](topleft),HαingreenandwarmH2moleculargasinblue(topmiddle),[Sivii](top
right),andPaα(bottomleft).Thelowest[OIII],Hα,H2,andPaαcontourscorrespondto5σ,whilefor[Sivii]thelowestcontourcorrespondsto2σ(to
bettermatchthemapofPrietoetal.2004).TheFoVis4arcsec×4arcsec.ThebrighterregionsintheKs/F814Wimagesdenotehigherextinctionandhence
dustabsorption.TheangularresolutionoftheKs/F814Wimagesis0.16arcsec,thatofH2is0.16arcsec(Mezcuaetal.2015)andthatof[Sivii]is0.19arcsec
(Prietoetal.2005).Bottommiddle:ChandraX-rayimageinthe0.5-7keVbandwithaFoVof8arcsec×8arcsec.Bottomright.ATCAradiocontinuum
mapat22GHz.ThesynthesizedbeamhasaFWHMof0.37arcsec×0.24arcsecataP.A.of-8deg.Contoursareplottedas(5,10,20,40,80)timesthe
off-sourcermsof3.3mJybeam−1.NorthisupandEastistotheleft.
(cid:13)c 2013RAS,MNRAS000,1–5
4 Mezcuaetal.
bythehorn-shapeddustlanesindicatesthattheionizingradiation Seyfert2likeCircinus.Athickerabsorberatthecentreistherefore
is not necessarily collimated by a nuclear torus but that dust at neededtocompletelyobscurethenucleusandproducetheobserved
scalesofafewtensofpcisresponsiblefortheobservedcone-like spectral energy distribution (SED). Still, as indicated before, our
morphology.Ifthisisthecase,amoreextendeddistributionshould A estimateisaconservativelowerlimit;thedustfilamentsmay
V
beobservedinHαwhichislessaffectedbydustextinction,asit get thicker as we approach the centre, in particular if they trace
isindeedthecase.Anevenmoreisotropicdistributionisexpected material flowing towards the centre. We see compelling evidence
and indeed observed when the ionized gas is traced at the same forthisflowofmaterialthroughdustfilamentsandlanesinother
∼30pcscalebymuchredderemissionlinesthatarenotsomuch nearby AGN (Prieto et al. 2014; Nadolny et al. in preparation).
affected by dust extinction (e.g. Paα or [Sivii] 2.48 µm). For Should that be the case, the dust filaments in Circinus must cast
example, the emission in [Sivii] is rather isotropic extending in higher A at the centre. The interferometric mapping of the
V
rather symmetric form in all directions (Fig. 2, top right). The centralpcindicatesalreadyastronggradientinthesilicate-derived
Paαemission(Fig.2,bottomleft)isevenmoreisotropicthanHα extinctionwithinthe2pc(theaverageextinctionagreeswithour
and[OIII],reinforcingthatthelatterarestronglyshapedbydust estimatefromtheSED):anincreasingextinctionisseenfromthe
extinction at a few tens of pc. Both the nuclear X-ray emission NWtotheSEclosetoaP.A.=-90deg(Tristrametal.2014),i.e.
(Fig. 2, bottom middle; see also Smith & Wilson 2001) and the inlinewiththeWesternhorn-shapedfilament.
pc-scale radio jet (Fig. 2, bottom right) also extend along the
directionsoftheionizedgas.TheX-rayemissionmaytentatively
3.4 Whereisthedustytorus?
be ascribed to the horn-shaped dust filaments. Still, the angular
resolution in these two spectral regions is insufficient to set up a We have seen that in the Circinus galaxy the collimation of the
preferentialdirectionormorphology. ionizingradiationdoesnotrequireapc-scaledustytorus.Instead,
alargerdustdistributionthatextendsupto∼30pcisresponsible
for collimating the ionizing radiation and hence for shaping the
3.3 Extinction
cone-like distribution of the ionized gas, which is by definition
To test to which degree the nuclear dust lanes contribute to the oneofthemainattributesofthetorus.Thislargerdustdistribution
obscurationofthenucleus,weestimatetheextinctionA produced manifestsitselfintheformofanuclearhorn-shapedstructureinour
V
by these dust lanes by measuring the galaxy colours in several extinctionmaps.Thisresult,togetherwiththeresultfrommid-IR
apertures(of0.1arcsecradius)inthedustystructureandcomparing interferometry that most of the pc-scale mid-IR dust emission
themtodust-freeregions(ofapertureradius0.1–0.2arcsec)located extends in the polar rather than in the perpendicular direction
atadistanceof∼3arcsec(∼60pc)fromthecentre.Wenotethat to the cone (Tristram et al. 2014), questions the presumed torus
thesedust-freeregionsmaystillbeaffectedbydustextinctionas (donut-like) morphology, at least at pc scales. A much higher
wearemeasuringatthelowestdetectionlevelofthedetector,thus obscurationthanthatinferredfromthehorn-shapedustfilaments
wearederivingalowerlimitontheA .Theextinctionlawfrom is still needed to explain the high levels of obscuration and the
V
Cardellietal.(1989)isapplied.Ontheassumptionofaforeground detection of broad emission lines in polarized light. The nucleus
dustscreen,whichisthemostappropriatedustdistributionforthe of Circinus at 2 and 5 µm resolves into a slightly elongated
filaments, an extinction of A (cid:38) 3 mag is obtained in the dusty source with a size of 1.9 ± 0.6 pc (Prieto et al. 2004). This
V
structure at a distance ∼1 arcsec (∼19 pc) from the nucleus and size is in line with that of the interferometric polar direction
A (cid:38)6.9magclosertothenucleus(atadistanceof0.2arcsec,∼4 component measured at 10 µm (Tristram et al. 2014) and is also
V
pc).Thisvalueisinagreementwithpreviousextinctionestimates, inbroadagreementwithourmeasuredshiftof3±0.6pcbetween
also assuming a foreground dust screen: close to the nucleus, a the optical and the IR peaks (Section 3.1). This shift could be
lowerlimitofA =6isderivedfromJ-Kextinctionmaps(Prieto interpretedasthesizeofthenuclearobscurer.Thedirectionofthe
V
et al. 2004); A = 6.3 mag is measured within the central 1.5 shift is coincident with that defined by the interferometric polar
V
arcsecofthestellarcomponentbyMaiolinoetal.(1998).Itshould dust component, but the outer region of the absorber may have
be noted that with this method we cannot measure the extinction escaped detection by the interferometric observations if it gets
directly on the nucleus as the colours at the nucleus locations progressivelylessopticallythickoutwards.Tristrametal.(2014)
andimmediatesurroundingrelatetodustemittingatabout2µm interpret the polar dust component as the true cause of Circinus
heated by the AGN. The location of our A measurements is a nuclearobscurationandcallitthetorus.Webroadlyagreewiththe
V
compromise between being the nearest to the centre and still far interpretation,butassociatetheobscurerwiththeinnermostparts
forthedusttobeheatedattherighttemperaturestoemitat2µm; ofthelarge-scale(∼30pc)dustfilamentscrossingthenucleus,the
therefore,theI-Kcoloursshouldreflectjustdustextinction. horn shape, for the following reasons: 1) the filaments shape the
Taking A (cid:38) 6.9 mag as a face value, we check next morphology of the ionization cone (Section 3.2); 2) their width,
V
whetherthisamountofextinctionisenoughtohidethenucleusof ∼0.2arcsec(4pc;Fig. 2),justfitintherangeoftheabsorbersize
Circinusintheoptical.Tothataim,wescaletheSeyfert1template discussed above; 3) they could naturally account for the nuclear
from Prieto et al. (2010), also based on subarcsec-resolution obscuration if getting thicker toward the inner parsecs, which is
observations,tothemid-IRfluxofCircinusmeasuredinthesame presumablythecasegiventhestrongextinctiongradientmeasured
work (17.6Jy at 18.72µm). Assuming a foreground extinction by the interferometric observations. In addition, no temperature
layer, we estimate that an A ∼ 20mag would be needed in gradient is observed down to 0.1 pc (Tristram et al. 2014),
V
order to obscure the Seyfert 1 template below the upper limit of which can be easily accommodated with a large-scale obscurer,
1.6mJyintheJ-bandderivedbyPrietoetal.(2010),inagreement filament-like,insteadofatorus.Still,thespatiallyresolvednuclear
with the silicate optical depth from Tristram et al. (2007) and in structureseenat2µm,of2pc,indicatesthatsomefractionofthe
line with the lower limit of A = 27.2 found by Burtscher et al. obscurer gets heated up to high temperatures as approaching the
V
(2015).Thus,thedusttracedinourmapsisabletoexplain(cid:38)1/3 centre.
of the extinction needed to turn a bright Seyfert 1 nucleus into a The results showed in this Letter therefore explain how an
(cid:13)c 2013RAS,MNRAS000,1–5
Circinusdustlane 5
ionizationconeofafewparsecsinsizecanbeshapedbythedust PrietoM.A.,MezcuaM.,Ferna´ndez-OntiverosJ.A.,Schartmann
distribution of several tens of parsec of the host galaxy, instead M.,2014,MNRAS,442,2145
ofapc-scalegeometrically-thicktoroidalstructure,demonstrating Prieto M. A., Reunanen J., Tristram K. R. W., Neumayer N.,
that the nuclear obscuration responsible for the type 1/2 AGN Fernandez-OntiverosJ.A.,OrientiM.,MeisenheimerK.,2010,
dichotomydoesnotnecessarilyarisefromthesamestructurethat MNRAS,402,724
collimates the ionizing radiation. If confirmed for a significantly ReunanenJ.,PrietoM.A.,SiebenmorgenR.,2010,MNRAS,402,
largersample-assuggestedbyadditionalcaseslikeNGC3169or 879
NGC7582(Prietoetal.2014)-thismightimplyarevisionofthe RisalitiG.,MaiolinoR.,SalvatiM.,1999,ApJ,522,157
dustdistributioninthetorus. Rousset G. et al., 2003, in Wizinowich P. L., Bonaccini
D., eds, Society of Photo-Optical Instrumentation Engineers
(SPIE) Conference Series Vol. 4839, Society of Photo-Optical
Instrumentation Engineers (SPIE) Conference Series. pp
ACKNOWLEDGEMENTS
140–149
The authors thank the referee, L. Burtscher, for his valuable SmithD.A.,WilsonA.S.,2001,ApJ,557,180
comments which have helped improve this manuscript, and M. TadhunterC.,TsvetanovZ.,1989,Nature,341,422
Elvisforinsightfuldiscussion.MMacknowledgessupportfromthe Tristram K. R. W., Burtscher L., Jaffe W., Meisenheimer K.,
NASAChandraGrantG05-16099X. Ho¨nig S. F., Kishimoto M., Schartmann M., Weigelt G., 2014,
A&A,563,A82
TristramK.R.W.etal.,2007,A&A,474,837
UrryC.M.,PadovaniP.,1995,PASP,107,803
REFERENCES
VeilleuxS.,Bland-HawthornJ.,1997,ApJ,479,L105
AntonucciR.,1993,ARA&A,31,473
AntonucciR.R.J.,MillerJ.S.,1985,ApJ,297,621
BianchiS.etal.,2012,MNRAS,426,3225
BurtscherL.etal.,2015,ArXiv:1511.05566
BurtscherL.etal.,2013,A&A,558,A149
CardelliJ.A.,ClaytonG.C.,MathisJ.S.,1989,ApJ,345,245
ElmouttieM.,HaynesR.F.,JonesK.L.,SadlerE.M.,EhleM.,
1998,MNRAS,297,1202
FreemanK.C.,KarlssonB.,LyngaG.,BurrellJ.F.,vanWoerden
H.,GossW.M.,MeboldU.,1977,A&A,55,445
Ho¨nig S. F., Gandhi P., Asmus D., Mushotzky R. F., Antonucci
R.,UedaY.,IchikawaK.,2014,MNRAS,438,647
Ho¨nig S. F., Kishimoto M., Antonucci R., Marconi A., Prieto
M.A.,TristramK.,WeigeltG.,2012,ApJ,755,149
Ho¨nig S. F., Kishimoto M., Gandhi P., Smette A., Asmus D.,
DuschlW.,PollettaM.,WeigeltG.,2010,A&A,515,A23
Ho¨nigS.F.etal.,2013,ApJ,771,87
JaffeW.etal.,2004,Nature,429,47
KayL.E.,1994,ApJ,430,196
Krolik J. H., 1998, Active Galactic Nuclei: From the Central
BlackHoletotheGalacticEnvironment
KrolikJ.H.,BegelmanM.C.,1988,ApJ,329,702
Lo´pez-Gonzaga N., Jaffe W., Burtscher L., Tristram K. R. W.,
MeisenheimerK.,2014,A&A,565,A71
Maiolino R., Krabbe A., Thatte N., Genzel R., 1998, ApJ, 493,
650
MattG.etal.,1999,A&A,341,L39
McConnellN.J.,MaC.-P.,2013,ApJ,764,184
MezcuaM.,PrietoM.A.,Ferna´ndez-OntiverosJ.A.,TristramK.,
NeumayerN.,KotilainenJ.K.,2015,MNRAS,452,4128
Oliva E., Marconi A., Cimatti A., di Serego Alighieri S., 1998,
A&A,329,L21
Oliva E., Salvati M., Moorwood A. F. M., Marconi A., 1994,
A&A,288,457
PackhamC.,RadomskiJ.T.,RocheP.F.,AitkenD.K.,Perlman
E.,Alonso-HerreroA.,ColinaL.,TelescoC.M.,2005,ApJ,618,
L17
PanessaF.etal.,2009,MNRAS,398,1951
PrietoM.A.,FreudlingW.,1993,ApJ,418,668
PrietoM.A.,MarcoO.,GallimoreJ.,2005,MNRAS,364,L28
PrietoM.A.etal.,2004,ApJ,614,135
(cid:13)c 2013RAS,MNRAS000,1–5