The predominance of infrared emission from

the polar region of accreting supermassive

black holes

Daniel Asmus!
ESO, Chile!
!
with!

Sebastian Hnig & Poshak Gandhi!

AGN are bright(est) in the mid-infrared.

Type 1
AGN

Type 2
AGN

Type 1
AGN

Type 2
AGN

???
!

WISE (0.4m)

WISE (0.4m)

VLT/VISIR (8.2m)

Asmus et al. 2014

Only few without strong star formation are resolved (~10-100pc)

Is the resolved emission coming from the torus?

Establishing a system axis from ionisation cones [OIII],

radio jets, maser disks, and polarized emission.

Circinus;
Wilson et. al. 2000

T
NGC 2110; credit: D. Evans

Circinus; Greenhill et al. 2003

The resolved emission is coming from the polar axis of the

AGN systems!

Angular difference (System Axis - MIR extension)

Asmus et al. 2016 (see also Braatz et al. 1993; Cameron et al. 1993; Bock et al. 2000;
Radomski et al. 2002, 2003; Whysong & Antonucci 2004; Packham et al. 2005;
Reunanen, Prieto & Siebenmorgen 2010; Hnig et al. 2010)

Is the mid-infrared emission of AGN dominated by dust

in/along the ionisation cone instead of the obscuring torus?

Relative amount of resolved emission

Asmus et al. 2016

Relative amount of resolved emission

The resolved emission strongly correlates with the [OIV]

emission produced in the ionisation cone

[OIV] emission
Asmus et al. 2016

NGC424; Hnig et al. 2012

NGC3783; Hnig et al. 2013

NGC1068;
Lpez Gonzaga et al.2014

Circinus;
Tristram et al. 2014

Polar elongation is dominant also on parsec scale

as found with MIDI interferometry (Lopez-Gonzaga et al. 2016)

The small and large scale elongation are aligned and seem to
trace the edge of the ionisation cone.

Circinus; Tristram+14

A new paradigm for the AGN dust structure?

Hnig et al. 2012

The mid-infraredX-ray correlation for all types of AGN is driven

by dust in the ionisation cones rather than the torus
Asmus et al. 2015!

46

(see also:!
Glass et al. 1982;!
Krabbe et al. 2001; !
Lutz et al. 2004; !
Horst et al. 2006; !
Ramos Almeida et al. 2007; !
Horst et al. 2008; !
Gandhi et al. 2009; !
Levenson et al. 2009; !
Fiore et al. 2009; !
Hardcastle et al. 2009; !
Hoenig et al. 2010; !
Asmus et al. 2011; !
Mullaney et al. 2011; !
Mason et al. 2012; !
Matsuta et al. 2012; !
Ichikawa et al. 2012; !
Sazonov et al. 2012;!
Mateos et al. 2015;!
Stern 2015)

44
log Lnuc(12m) [erg/s]

Observed MIR luminosity

45

43

42

41

40

39
39

40

41
42
43
44
int
log L (2-10keV) [erg/s]

45

Absorption-corr. X-ray luminosity

46

What have we learned?

High angular resolution is

important for AGN studies
The AGN structures remain
mostly unresolved in singledish observations
The resolved emission is
coming from dust in/along the
ionisation cone
The mid-infrared emission of
AGN is probably dominated by
the polar emission (instead of
the torus)

Back-up Slides

The chop & nod technique makes the target observable.

The resolved emission is not correlated to the host orientation.

Angular difference (System Axis - MIR extension)

Asmus et al. 2016

The outliers

Seyfert 1.2
!

[OIII] pointlike
!

marginally
resolved
!

PA error 27

Seyfert 1.5/2
!

wide opening
ionisation cones
!

edge-on spiral
!

emission traces
host dust lane

Seyfert 1.2
!

wide opening
ionisation cones
!

weakest AGN in
the sample
!

low S/N

The reasons for the low detection rates

Distance

low S/N

unresolved

intrinsic weakness

Inclination

no elongation

Difference between type I and II is smaller than expected

46

1.0

fraction of total

0.8

45

0.4
0.2
20

43

42

Sy 1-1.5 (>=3 epochs)

Sy 1.8-1.9 (>=3 epochs)
Sy 2 (>=3 epochs)
LINER (>=3 epochs)
Sy 1-1.5 (<3 epochs)
Sy 1.8-1.9 (<3 epochs)
Sy 2 (<3 epochs)
LINER (<3 epochs)
NLS1
Compton-thick

41

40

39
39

number of objects

log Lnuc(12m) [erg/s]

44

0.6

40

41

42
43
44
log L (2-10keV) [erg/s]
int

45

15

10

0
-0.6

46

-0.3

0.0
0.3
0.6
0.9
1.2
1.5
nuc
int
log L (12m) - log L (2-10keV)

1.8