Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
Molecular torus in the radio galaxy NGC 1052Seiji Kameno (NAOJ / Joint ALMA Observatory)V. Impellizzeri, D. Espada, S. Martin, S. Sawada-Satoh, N. Nakai, H. Sugai, Y. Terashima, K. Kohno, L. Minju
collaborated with
Fernández-Ontiveros+2011Ravindranath+2001
10˝=1 kpc
/ 18ALMA EA Science Workshp 2017 in Daejeon, Seiji Kameno : Molecular torus in the radio galaxy NGC 1052 2
Artist’s impression from NASA site
• Energy source
• Evolution of SMBH
What controls the accretion rate?
Mass accretion in AGN
/ 18ALMA EA Science Workshp 2017 in Daejeon, Seiji Kameno : Molecular torus in the radio galaxy NGC 1052
Mass accretion from galactic disk into the center
3
What is happening in the transition region (edge of SoI)?
• M(R) ~ MBH
• 1 pc < R < 10 pc • Change in the rotation curve • Change in the rotation axis
Black-hole gravisphere galactic disk
rotation curve
SMBH gravisphere
TransitionRegion
0.2 0.5 1.0 2.0 5.0 20.0 50.0 200.0
1e
+0
11
e+
02
1e
+0
31
e+
04
1e
+0
5
Radius [pc]
Sp
eci
fic A
ng
ula
r M
om
em
tum
[p
c km
/s]
Circinus Galaxy
Keplerian
Galactic R
otation
angular momentum
The Transition Region betweenSMBH gravisphere (a.k.a. SoI)and galactic disk
Sofue & Rubin (200
• VLBI resolution is crucial. • Now ALMA is accessible there!
~ 0.1 arcsec in nearby AGNs
radius
rotat
ion ve
locity
/ 18ALMA EA Science Workshp 2017 in Daejeon, Seiji Kameno : Molecular torus in the radio galaxy NGC 1052
Parsec-scale mass accretion mechanism in AGN
4
Question: What is the key mechanism of angular momentum transfer?
What about non-starburst AGNs?
• Turbulent gas? • Radiation drag? • Dynamical friction of star clusters?
Starburst-AGN connection• nuclear starburst • turbulence by supernovae
Simulation: Wada+09, ApJ, 702, 63
ALMA observations of NGC 1097: Fathi+13, ApJL, 770, L23
ALMA observations toward NGC 1052
Host galaxy E4 BT=11.41 mag Distance 20.3 Mpc, 1˝=98 pc Velocity Vsys(LSR, Radio) = 1471 km s-1 Radio continuum 0.4 Jy@345 GHz Radio Jet β = 0.25, i=62º±10º H2O maser velocity = 1400 - 1850 km/s Line absorption HI, OH, HCO+, HCN, CO Free-free absorption
1 pc
(c)
(d)
Good target to observe mass accretion in a molecular torus
composed of cold dense plasma, molecular gas, and the XDRbetween them, might control the continuum and maser fluxdensities. Cold dense plasma in the torus attenuates the con-tinuum emission from the jet. Since free-free absorption (FFA)is substantial along the western jet within 10 mas of the centralengine, the time variability of the continuum flux density can beascribed to changes of the attenuation through cold denseplasma. When a knot runs behind the inhomogeneous mixtureof plasma andmolecular gas in the torus, the attenuation by FFAand the maser gain along the line of sight would change. Escapefrom the ionized region will increase the continuum flux den-sity and entering the molecular cloud increases the maser fluxdensity and decreases the maser velocity width at the same time(see Fig. 3). The optical depth in terms of FFA is a function offrequency ! in GHz as
"! ¼ "0!"2:1; ð5Þ
and the FFA coefficient "0 in NGC 1052 can be regressed as"0 ¼ 245 #"1, where # is the separation from the central enginein milliarcseconds (Kameno et al. 2003). The gradient of the op-
tical depth, d" /d#, at 22 GHz and at # ¼ 0:31 mas can pro-duce a decrease of attenuation by 20% and an apparent motionof 0.03 mas for 0.25c in 3 days.
4.3.3. Interactingg Maser Model
Interacting masers are another mechanism for extremelybright masers. When two maser clouds sized %1014 cm arealigned along the line of sight (‘‘interaction’’) with a separationof%1016 cm, the maser emission is beamed and high brightnesstemperatures up to 1017 K can be produced (Deguchi &Watson1989). The synchronicity of continuum and maser flares is notrequired in this model. The typical relative velocity of twoclouds in NGC 1052 can be estimated as %200 km s"1 fromthe velocity width of a broad maser. Motion in %106 s will be%1011 cm, much less than the typical size of a maser cloud.Thus, it is difficult to change the alignment of two maser cloudsin such a short timescale.
4.3.4. The Jet Maser Model
As has been explained for the flares of maser and contin-uum in Mrk 348 (Peck et al. 2003), an interaction between theradio jet and an encroaching molecular gas can produce a
Fig. 3.—Schematic diagram of the jets and torus in NGC 1052. Double-sided jets are inclined by k57& with respect to the line of sight. The cross section of ageometrically thick molecular torus perpendicular to the jet is drawn. The inner face of the torus is ionized by irradiation from the central engine and causes free-freeabsorption. The X-ray dissociation region (XDR) is formed in the interlayer, where excited molecular gas performs maser amplification.
NARROW H2O MASER FEATURE IN NGC 1052 149No. 1, 2005
Sawada-Satoh+2008 Kameno+2003
Kameno+2005
/ 18ALMA EA Science Workshp 2017 in Daejeon, Seiji Kameno : Molecular torus in the radio galaxy NGC 1052
Molecular gas distribution and velocity (CO)
6
CO (3-2) total intensity map
core
SMG
jet
moment-1 (velocity field)
Vsys→
/ 18ALMA EA Science Workshp 2017 in Daejeon, Seiji Kameno : Molecular torus in the radio galaxy NGC 1052
CND (CircumNuclear disk) rotation
7
moment-1
V
conti
nuum
• radius ~ 100 pc • rotation speed ~ 150 km s-1 • enclosed mass = 5x108 M⊙
/ 18ALMA EA Science Workshp 2017 in Daejeon, Seiji Kameno : Molecular torus in the radio galaxy NGC 1052
Long-baseline view
8
000.26⇥ 000.20 beam
000.065⇥ 000.053 beam
/ 18ALMA EA Science Workshp 2017 in Daejeon, Seiji Kameno : Molecular torus in the radio galaxy NGC 1052
High-resolution velocity field
9
←Vsys
different components from the CND
/ 18ALMA EA Science Workshp 2017 in Daejeon, Seiji Kameno : Molecular torus in the radio galaxy NGC 1052
Angular momentum plot
10
SMBH gravisphere
Transition
Region
0.2 0.5 1.0 2.0 5.0 20.0 50.0 200.0
1e+
01
1e+
02
1e+
03
1e+
04
1e+
05
Radius [pc]
Speci
fic A
ngula
r M
om
em
tum
[pc
km/s
] Circinus Galaxy ● and NGC 1052 ●
Keplerian
Galactic Rotation
• Still outside of gravisphere • Need higher resolutions!
Absorption Studies with ALMA and KVN
/ 18ALMA EA Science Workshp 2017 in Daejeon, Seiji Kameno : Molecular torus in the radio galaxy NGC 1052
CO line profile toward the nucleus
12
Spectra towardthe nucleus
CO (J=2-1) : inverse P-Cyg profileblue emission + red absorption implication of inward gas motion
/ 18ALMA EA Science Workshp 2017 in Daejeon, Seiji Kameno : Molecular torus in the radio galaxy NGC 1052
Absorption features
13
Spectra towardthe nucleus
338 339 340 341 342 343 344 345
0.32
0.34
0.36
0.38
0.40
0.42
NGC 1052 Cycle 2 Band 7
Frequency [GHz]
Flux D
ensity
[Jy]
J0301+0118 (x1.9)
J0334−4008 (x0.78)
J0219+0121 (x2.25)
J2357−5311 (x0.68)
−100
Vsys
100
H13CN v=0 J=4−3
−100
Vsys
100
200
CO J=3−2
−100
Vsys
100
200
CS J=7−6 v=0
500
600
700
800
900
1000
SiO v=0 J=8−7
200
300
400
500
600
700
800
900
1000
SiO J=8−7 v=1
500
600
700
800
900
1000
SiO J=8−7 v=2
−100
Vsys
100
200
CN N=3−2 J=5/2−3/2
−100
Vsys
100
200
CN N=3−2 J=7/2−5/2
−100
Vsys
100
SO J_N=8_8−7_7
−100
Vsys
100
SO J_N=7_8−6_7
−100
Vsys
100
HC15N v=0 J=4−3
−100
Vsys
100
HCS+ J=8−7
−100
Vsys
100
HC3N J=38−37
351 352 353 354 355 356 357 358
0.32
0.34
0.36
0.38
0.40
0.42
0.44
NGC 1052 Cycle 2 Band 7
Frequency [GHz]
Flux D
ensity
[Jy]
J0301+0118 (x1.9)
J0334−4008 (x0.78)
J0219+0121 (x2,25)
J2357−5311 (x0.68)
−100
Vsys
100
HCN J=4−3 v=0
−100
Vsys
100
HCN J=4−3 v2=1
−100
Vsys
100
HC3N J=39−38
−100
Vsys
100
HCO+ J=4−3
−100
Vsys
100
HCN v2=2 J=4−3 l=2f
−100
Vsys
100
HCN v2=2 J=4−3 l=2e
−100
Vsys
100
HCN v2=1 J=4−3 l=1f
−100
Vsys
100
HCN v2=2 J=4−3 l=0
−100
Vsys
100
HCO+ v2=1 J=4−3 l=1e
−100
Vsys
100
HCO+ v2=1 J=4−3 l=1f
−100
Vsys
100
HCN v2=3 J=4−3 l=1f
228.5 229.0 229.5 230.0 230.5 231.0 231.5
0.42
0.44
0.46
0.48
0.50
NGC 1052 Cycle 2 Band 6
Frequency [GHz]
Flux D
ensity
[Jy]
J0238+1636 (x0.16)
J0239−0234 (x2.26)
−200
−100
Vsys
100
200
300
400
CO J=2−1
−100
Vsys
100
200
300
400
500
600
700
800
900
1000
1100
1200
1300
H30a
−200
−100
Vsys
100
200
300
400
H2O 5(5,0)−6(4,3) v2=1
−100
Vsys
100
200
SO2 v=0 37(10,28)−38(9,29)
−100
Vsys
100
200
SO2 v2=1 14(3,11)−14(2,12)
213 214 215 216 217
0.48
0.49
0.50
0.51
0.52
NGC 1052 Cycle 2 Band 6
Frequency [GHz]
Flux D
ensity
[Jy]
J0238+1636 (x0.18)
J0239−0234 (x2.6)
−100
Vsys
100
200
SO 5_5−4_4
−100
Vsys
100
200
SO 7_8−7_7
Vsys
200
400
600
800
1000
1200
1400
1600
1800
SiO J=5−4 v=0
Vsys
200
400
600
800
1000
1200
1400
1600
1800
SiO J=5−4 v=1
−100
Vsys
100
200
SO2 v=0 22(2,20)−22(1,21)
−100
Vsys
100
200
SO2 v=0 16(3,13)−16(2,14)
−200
−100
Vsys
100
200
34SO v=0 6(5)−5(4)
−100
Vsys
100
200
H2S 2(2,0)−2(1,1)
CO, HCN, HCO+, CS, SO, and CN- isotopologues : H13CN, HC15N- vib-excited HCN, HCO+
/ 18ALMA EA Science Workshp 2017 in Daejeon, Seiji Kameno : Molecular torus in the radio galaxy NGC 1052
Optical depths
14
1300 1400 1500 1600 1700 1800
0.0
0.1
0.2
0.3
0.4
0.5
0.6
NGC 1052
LSR Velocity [km/s]
Opt
ical
Dep
th (
+ o
ffset
)
HCN J=4−3 v=0
CO J=3−2
HCO+ J=4−3
CO J=2−1
HCN v2=1 J=4−3 l=1f
H13CN v=0 J=4−3
SO 5_5−4_4
SO J_N=7_8−6_7
CS J=7−6 v=0
HCO+ v2=1 J=4−3 l=1f
Vsys
V sys+250�100 km s�1
HCN
HCN v2=1
H13CN
CO (J=3-2)
CO (J=2-1)
HCO+ v2=1
HCO+
SO
SO
CS
EW =
Z⌧(v) dv = 24.4 kms�1
• Mostly redshifted w.r.t. Vsys • Wider than CNS • HCN deeper than CO
Absorption features are likely tooriginate in a molecular torus
/ 18ALMA EA Science Workshp 2017 in Daejeon, Seiji Kameno : Molecular torus in the radio galaxy NGC 1052
Presence of vibrationally excited lines
15
1300 1400 1500 1600 1700 1800
0.0
0.1
0.2
0.3
0.4
0.5
0.6
NGC 1052
LSR Velocity [km/s]
Opt
ical
Dep
th (
+ o
ffset
)
HCN J=4−3 v=0
CO J=3−2
HCO+ J=4−3
CO J=2−1
HCN v2=1 J=4−3 l=1f
H13CN v=0 J=4−3
SO 5_5−4_4
SO J_N=7_8−6_7
CS J=7−6 v=0
HCO+ v2=1 J=4−3 l=1f
Vsys
HCN
HCN v2=1
H13CN
CO (3-2)
CO (2-1)
HCO+ v2=1
HCO+
SOSO
CS
HCN J=4-3 and HCO+ J=4-3• line ratio (v=0 to v2=1) : R=0.6 • if optically thin, Tex = 520 K • IR (14 µm) pumping from hot dust?
Sakamoto+2010, ApJL, 725, L228
/ 18ALMA EA Science Workshp 2017 in Daejeon, Seiji Kameno : Molecular torus in the radio galaxy NGC 1052
Locating HCN absorption with KVN
16
millimeter and submillimeter interferometric observations havedetected the vibrationally excited HCN emission lines in thedust-enshrouded AGN of the luminous infrared galaxies(Sakamoto et al. 2010; Imanishi & Nakanishi 2013). Thederived H2 column density (NH2) ranges 10
24–1025 cm−2, usingthe values of NHCN in Table 1. Hence, equivalent NH would beof the same order as NH2. It is one or two orders higher than theestimate of the NH∼1023 cm−2 from modeling of X-rayspectra (Guainazzi & Antonelli 1999; Weaver et al. 1999;Guainazzi et al. 2000; Brenneman et al. 2009), and the electroncolumn density of ∼1023 cm−2 estimated from the FFA opacityof the dense plasma (Sawada-Satoh et al. 2009).
4.2. Circumnuclear Torus Model
Kameno et al. (2005) and Sawada-Satoh et al. (2008)proposed that the circumnuclear torus consists of several phaselayers. On the inner surface of the torus, a hot (∼8000 K)plasma layer is formed by X-ray emission from the centralengine and causes FFA. A heated (above ∼400 K) molecularlayer or X-ray dissociation region (XDR; Neufeld et al. 1994;Neufeld & Maloney 1995) lies next to the plasma layer, and theH2O megamaser emission arises from here. As well as FFAopacity distribution, the HCN opacity shows high values on thewestern receding side of the jet. Thus, HCN absorbing gascould be associated with the torus, like the dense plasma. If Texof HCN is ∼230 K, HCN molecules should lie in the coolermolecule layer next to XDR. The redshifted HCN absorptionagainst the continuum emission is likely indicative of ongoinggas infall. The velocities of the HCN absorption features areclose to the centroid velocity of the broad H2O maser emission(∼1700 km s−1). Therefore, H2O and HCN could trace thesame infall motion inside the torus. The HCN absorptionspectrum consists of at least two narrow absorption features,and HCN absorbing gas is more likely to be several small(�0.1 pc) clumps or layers with a different velocity, rather thana large homogeneous structure with a single velocity. We haveseen the opacity contrast between the absorption features I andII, and the contrast also suggests inhomogeneity inside themolecular torus. In Figure 3, we present a possible model forthe geometry of the circumnuclear torus and the jet in NGC
1052. Since the jet axis is inclined from the sky plane, the nearside of the torus should cover the receding jet side. The line ofsight intersects at least two HCN gas clumps inside the torus.As we have mentioned in Section 3, there is no significant
detection of absorption features around 1500–1600 km s−1,close to Vsys. A possible explanation is that the missingabsorption features could arise not from the compact molecularclumps in the circumnuclear region, but from the foregrounddiffuse interstellar medium in the host galaxy. If so, theircovering factor would not vary on scales between a few parsecsand a few hundred parsecs, and thus the depth of the absorptionfeatures would be ∼−3%, the same as the PdBI results. It iscomparable to the rms noise level of our spectral profile withthe KVN, and no significant detection comes as a consequence.Furthermore, when their background continuum sources have afaint and extended structure, the background continuumemissions should be fully resolved-out. Therefore, the absorp-tion features against the resolved-out background sourceswould be invisible in our KVN data.
Figure 1. Spectral profile of the HCN (1–0) absorption line toward the centerof NGC1052 obtained with the KVN. Two velocity features of HCNabsorption were detected at 1656 and 1719 km s−1. We gave labels to theformer (I) and the latter (II), respectively. The rms noise in the normalized fluxdensity is 0.035.
Figure 2. Color images of the HCN (1–0) optical depth of (a) feature I (1656km s−1) and (b) feature II (1719 km s−1) in the circumnuclear region ofNGC1052, overlaid by a contour map of 89 GHz continuum emission. Thecontour starts at the 3σ level, increasing by a factor of 2, where σ = 3.9mJybeam−1. The absorption maps have achieved a 1σ noise level of 4.6 mJybeam−1 in intensity, or ∼0.02 in optical depth. The velocity width of onechannel map is 52.9 km s−1.
3
The Astrophysical Journal Letters, 830:L3 (5pp), 2016 October 10 Sawada-Satoh et al.
• absorption feature toward receding jet
• clumpy, with a filling factor ~ 0.03
millimeter and submillimeter interferometric observations havedetected the vibrationally excited HCN emission lines in thedust-enshrouded AGN of the luminous infrared galaxies(Sakamoto et al. 2010; Imanishi & Nakanishi 2013). Thederived H2 column density (NH2) ranges 10
24–1025 cm−2, usingthe values of NHCN in Table 1. Hence, equivalent NH would beof the same order as NH2. It is one or two orders higher than theestimate of the NH∼1023 cm−2 from modeling of X-rayspectra (Guainazzi & Antonelli 1999; Weaver et al. 1999;Guainazzi et al. 2000; Brenneman et al. 2009), and the electroncolumn density of ∼1023 cm−2 estimated from the FFA opacityof the dense plasma (Sawada-Satoh et al. 2009).
4.2. Circumnuclear Torus Model
Kameno et al. (2005) and Sawada-Satoh et al. (2008)proposed that the circumnuclear torus consists of several phaselayers. On the inner surface of the torus, a hot (∼8000 K)plasma layer is formed by X-ray emission from the centralengine and causes FFA. A heated (above ∼400 K) molecularlayer or X-ray dissociation region (XDR; Neufeld et al. 1994;Neufeld & Maloney 1995) lies next to the plasma layer, and theH2O megamaser emission arises from here. As well as FFAopacity distribution, the HCN opacity shows high values on thewestern receding side of the jet. Thus, HCN absorbing gascould be associated with the torus, like the dense plasma. If Texof HCN is ∼230 K, HCN molecules should lie in the coolermolecule layer next to XDR. The redshifted HCN absorptionagainst the continuum emission is likely indicative of ongoinggas infall. The velocities of the HCN absorption features areclose to the centroid velocity of the broad H2O maser emission(∼1700 km s−1). Therefore, H2O and HCN could trace thesame infall motion inside the torus. The HCN absorptionspectrum consists of at least two narrow absorption features,and HCN absorbing gas is more likely to be several small(�0.1 pc) clumps or layers with a different velocity, rather thana large homogeneous structure with a single velocity. We haveseen the opacity contrast between the absorption features I andII, and the contrast also suggests inhomogeneity inside themolecular torus. In Figure 3, we present a possible model forthe geometry of the circumnuclear torus and the jet in NGC
1052. Since the jet axis is inclined from the sky plane, the nearside of the torus should cover the receding jet side. The line ofsight intersects at least two HCN gas clumps inside the torus.As we have mentioned in Section 3, there is no significant
detection of absorption features around 1500–1600 km s−1,close to Vsys. A possible explanation is that the missingabsorption features could arise not from the compact molecularclumps in the circumnuclear region, but from the foregrounddiffuse interstellar medium in the host galaxy. If so, theircovering factor would not vary on scales between a few parsecsand a few hundred parsecs, and thus the depth of the absorptionfeatures would be ∼−3%, the same as the PdBI results. It iscomparable to the rms noise level of our spectral profile withthe KVN, and no significant detection comes as a consequence.Furthermore, when their background continuum sources have afaint and extended structure, the background continuumemissions should be fully resolved-out. Therefore, the absorp-tion features against the resolved-out background sourceswould be invisible in our KVN data.
Figure 1. Spectral profile of the HCN (1–0) absorption line toward the centerof NGC1052 obtained with the KVN. Two velocity features of HCNabsorption were detected at 1656 and 1719 km s−1. We gave labels to theformer (I) and the latter (II), respectively. The rms noise in the normalized fluxdensity is 0.035.
Figure 2. Color images of the HCN (1–0) optical depth of (a) feature I (1656km s−1) and (b) feature II (1719 km s−1) in the circumnuclear region ofNGC1052, overlaid by a contour map of 89 GHz continuum emission. Thecontour starts at the 3σ level, increasing by a factor of 2, where σ = 3.9mJybeam−1. The absorption maps have achieved a 1σ noise level of 4.6 mJybeam−1 in intensity, or ∼0.02 in optical depth. The velocity width of onechannel map is 52.9 km s−1.
3
The Astrophysical Journal Letters, 830:L3 (5pp), 2016 October 10 Sawada-Satoh et al.
Sawada-Satoh+2016: HCN absorption with KVN
We can estimate an approximate size of the foregroundabsorbing molecular gas using the relation =N f n LvH H2 2 , wherefv is the volume filling factor, nH2 is the molecular hydrogenvolume density, and L is the size of the molecular gas. Here wesimply assume fv = 0.01. Adopting NH2 of 10
24–1025 cm−2 fromour results and nH2=(0.1–2)×106 cm−3 in the Galactic CND(Smith & Wardle 2014), the resulting L is approximately on 1-pcscales. It is consistent with the approximate size of thecircumnuclear torus of NGC1052 (Kameno et al. 2001). Onthe other hand, the size of the innermost receding jet componentcan be estimated as 0.2–0.4 mas from the VLBA images at43GHz (Kadler et al. 2004; Sawada-Satoh et al. 2008), and itcorresponds to 0.02–0.04 pc in linear scale.
Based on our measurements of the radial velocities and thedistribution of HCN absorption features, the mass infall rate ofthe gas accretion toward the central engine can be estimated by
r= WM f R Vv in2
in˙ , where fv is the volume filling factor, Rin is aninfall radius, ρ is the mass density of the infalling materials, Vinis the infall velocity, and Ω is the solid angle of the torus fromthe center. Here we assume r = N m RH H in, where NH is thecolumn density of a hydrogen atom and mH is the mass of ahydrogen atom. Giving Rin = 1 pc, NH=1024–1025 cm−2, and
Vin = 200 km s−1 as the approximate velocity of the HCNabsorption feature with respect to Vsys ( -V Vp sys in Table 1),the derived mass infall rate ranges M=(47–470) fv(Ω/4π)M☉ yr−1. If (Ω/4π) takes a few 0.1 and fv is 0.01, it wouldbe comparable to the calculation of the accretion rate of10−1.39M☉ yr−1 from the hard X-ray luminosities by Wu &Cao (2006). We note that the X-ray luminosity indicates aninstantaneous accretion rate. However, our estimation of theaccretion rate in the molecular torus gives a long-term (Rin/Vin∼5000 year) activity. The coincidence of the two accretionrate values suggests the continuity of AGN activity inNGC1052.
5. SUMMARY
We have conducted 1 mas angular resolution observationstoward the HCN(1–0) absorption of the circumnuclar region inNGC1052 with the KVN. Two HCN absorption features areidentified and reach a depth of �10% from the continuum levelat 1656 and 1719 km s−1, redshifted with respect to Vsys. Wefind an NHCN of 1015–1016 cm−2 in the center of NGC1052,assuming an HCN covering factor of 1 and a Tex of 100–230 K.
Table 1Column Density of HCN (1–0) Absorption
Label Vp Vp − Vsys Δv ò tdv NHCN,100 NHCN,230(km s−1) (km s−1) (km s−1) (km s−1) (1014 cm−2) (1014 cm−2)
(1) (2) (3) (4) (5) (6) (7)
I 1656 149 31.7 9.3±0.6 9.5±0.6 50±3II 1719 212 52.9 19±1 20±1 101±5
Note. (1) Absorption feature ID, shown in Figure 1. (2) Peak velocity of the absorption feature. (3) Velocity with respect to Vsys. (4) Line width. (5) Velocity-integrated optical depth. (6) HCN column density with Tex=100 K. (7) HCN column density with Tex=230 K.
Figure 3. (a) Possible model of the oriented double-sided jet and the circumnuclear torus. The near side of the torus covers the receding jet side. (b) Schematic diagramfor the intersection of the circumnuclear torus in NGC1052. The torus has two-phase layers on the inner surface, the X-ray heated plasma region and the X-raydissociation region (XDR). HCN absorbing gas is located in the cooler molecule region next to XDR. HCN absorbing gas could be clumpy, and the line of sight passesthrough several clumpy clouds with a different radial velocity.
4
The Astrophysical Journal Letters, 830:L3 (5pp), 2016 October 10 Sawada-Satoh et al.
/ 18ALMA EA Science Workshp 2017 in Daejeon, Seiji Kameno : Molecular torus in the radio galaxy NGC 1052
Absorption features and H2O maser
17
1000 1200 1400 1600 1800 2000
0.00
0.02
0.04
0.06
NGC 1052
LSR Velocity [km/s]
Optic
al De
pth
● ● ●●
●
●●
●
●
●●
●●
●
●●
● ●
●●
●
●
● ●
●
●●
●
●
● ●●
● ●●
●
●
●
●
●●
●●
●
●
●●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
● ●
●●
●
●
●
●● ●
● ●● ●
● ●●
● ●
●●
●
● ● ●●
●●
●
●
●
●
●●
●●
●
●●
●●
●
●
●●●
●
●
●
●●
●●
●●
●
●●
●●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●●●
●
●●
●
●
●
●
●
●
●
●
●●
●
●
●
●●
●
●
●
●
●●●●
●
●
●
●●
●
●
●
●
●
●●
●
●
●●
●●
●●●
●
●
●
●●
●● ●
● ●
●● ●
● ● ●●
●●
●●
●
●●
● ● ●
●●
●
●
●●
● ●
● ●●
●
●
●●
●
●● ●
●
● ●●
●
● ●
●
● ●
● ● ● ●
●
●●
●
●
● ●
●
●
●●
●
●●
●
● ●
●
●
● ●●
● ●
●
●
●
●● ●
●
●●
●
●
● ●
●
●
●
●
● ●
●
● ● ●
●
●
● ●
● ●●
● ●●
●
●●
●
●
●●
●●
●
●
●●
●●
● ●●
●● ●
● ●
●
● ●●
●●
●
HCN J=1-0
SO JN=32-21
SO JN=33-22
1400 1600 1800 2000
0.00
0.05
0.10
0.15
NGC 1052 H2O maser
LSR Velocity [km/s]
Flux D
ensit
y [Jy
]
Vsys
• Asymmetric profile - sharp red edge
• Less redshifted than H2O maser
• Inside molecular torus?
millimeter and submillimeter interferometric observations havedetected the vibrationally excited HCN emission lines in thedust-enshrouded AGN of the luminous infrared galaxies(Sakamoto et al. 2010; Imanishi & Nakanishi 2013). Thederived H2 column density (NH2) ranges 10
24–1025 cm−2, usingthe values of NHCN in Table 1. Hence, equivalent NH would beof the same order as NH2. It is one or two orders higher than theestimate of the NH∼1023 cm−2 from modeling of X-rayspectra (Guainazzi & Antonelli 1999; Weaver et al. 1999;Guainazzi et al. 2000; Brenneman et al. 2009), and the electroncolumn density of ∼1023 cm−2 estimated from the FFA opacityof the dense plasma (Sawada-Satoh et al. 2009).
4.2. Circumnuclear Torus Model
Kameno et al. (2005) and Sawada-Satoh et al. (2008)proposed that the circumnuclear torus consists of several phaselayers. On the inner surface of the torus, a hot (∼8000 K)plasma layer is formed by X-ray emission from the centralengine and causes FFA. A heated (above ∼400 K) molecularlayer or X-ray dissociation region (XDR; Neufeld et al. 1994;Neufeld & Maloney 1995) lies next to the plasma layer, and theH2O megamaser emission arises from here. As well as FFAopacity distribution, the HCN opacity shows high values on thewestern receding side of the jet. Thus, HCN absorbing gascould be associated with the torus, like the dense plasma. If Texof HCN is ∼230 K, HCN molecules should lie in the coolermolecule layer next to XDR. The redshifted HCN absorptionagainst the continuum emission is likely indicative of ongoinggas infall. The velocities of the HCN absorption features areclose to the centroid velocity of the broad H2O maser emission(∼1700 km s−1). Therefore, H2O and HCN could trace thesame infall motion inside the torus. The HCN absorptionspectrum consists of at least two narrow absorption features,and HCN absorbing gas is more likely to be several small(�0.1 pc) clumps or layers with a different velocity, rather thana large homogeneous structure with a single velocity. We haveseen the opacity contrast between the absorption features I andII, and the contrast also suggests inhomogeneity inside themolecular torus. In Figure 3, we present a possible model forthe geometry of the circumnuclear torus and the jet in NGC
1052. Since the jet axis is inclined from the sky plane, the nearside of the torus should cover the receding jet side. The line ofsight intersects at least two HCN gas clumps inside the torus.As we have mentioned in Section 3, there is no significant
detection of absorption features around 1500–1600 km s−1,close to Vsys. A possible explanation is that the missingabsorption features could arise not from the compact molecularclumps in the circumnuclear region, but from the foregrounddiffuse interstellar medium in the host galaxy. If so, theircovering factor would not vary on scales between a few parsecsand a few hundred parsecs, and thus the depth of the absorptionfeatures would be ∼−3%, the same as the PdBI results. It iscomparable to the rms noise level of our spectral profile withthe KVN, and no significant detection comes as a consequence.Furthermore, when their background continuum sources have afaint and extended structure, the background continuumemissions should be fully resolved-out. Therefore, the absorp-tion features against the resolved-out background sourceswould be invisible in our KVN data.
Figure 1. Spectral profile of the HCN (1–0) absorption line toward the centerof NGC1052 obtained with the KVN. Two velocity features of HCNabsorption were detected at 1656 and 1719 km s−1. We gave labels to theformer (I) and the latter (II), respectively. The rms noise in the normalized fluxdensity is 0.035.
Figure 2. Color images of the HCN (1–0) optical depth of (a) feature I (1656km s−1) and (b) feature II (1719 km s−1) in the circumnuclear region ofNGC1052, overlaid by a contour map of 89 GHz continuum emission. Thecontour starts at the 3σ level, increasing by a factor of 2, where σ = 3.9mJybeam−1. The absorption maps have achieved a 1σ noise level of 4.6 mJybeam−1 in intensity, or ∼0.02 in optical depth. The velocity width of onechannel map is 52.9 km s−1.
3
The Astrophysical Journal Letters, 830:L3 (5pp), 2016 October 10 Sawada-Satoh et al.
1 pc
(c)
(d)
0.5 pc
0.5 pc
Vsys
HCNH2O maser
plasma free-free
composed of cold dense plasma, molecular gas, and the XDRbetween them, might control the continuum and maser fluxdensities. Cold dense plasma in the torus attenuates the con-tinuum emission from the jet. Since free-free absorption (FFA)is substantial along the western jet within 10 mas of the centralengine, the time variability of the continuum flux density can beascribed to changes of the attenuation through cold denseplasma. When a knot runs behind the inhomogeneous mixtureof plasma andmolecular gas in the torus, the attenuation by FFAand the maser gain along the line of sight would change. Escapefrom the ionized region will increase the continuum flux den-sity and entering the molecular cloud increases the maser fluxdensity and decreases the maser velocity width at the same time(see Fig. 3). The optical depth in terms of FFA is a function offrequency ! in GHz as
"! ¼ "0!"2:1; ð5Þ
and the FFA coefficient "0 in NGC 1052 can be regressed as"0 ¼ 245 #"1, where # is the separation from the central enginein milliarcseconds (Kameno et al. 2003). The gradient of the op-
tical depth, d" /d#, at 22 GHz and at # ¼ 0:31 mas can pro-duce a decrease of attenuation by 20% and an apparent motionof 0.03 mas for 0.25c in 3 days.
4.3.3. Interactingg Maser Model
Interacting masers are another mechanism for extremelybright masers. When two maser clouds sized %1014 cm arealigned along the line of sight (‘‘interaction’’) with a separationof%1016 cm, the maser emission is beamed and high brightnesstemperatures up to 1017 K can be produced (Deguchi &Watson1989). The synchronicity of continuum and maser flares is notrequired in this model. The typical relative velocity of twoclouds in NGC 1052 can be estimated as %200 km s"1 fromthe velocity width of a broad maser. Motion in %106 s will be%1011 cm, much less than the typical size of a maser cloud.Thus, it is difficult to change the alignment of two maser cloudsin such a short timescale.
4.3.4. The Jet Maser Model
As has been explained for the flares of maser and contin-uum in Mrk 348 (Peck et al. 2003), an interaction between theradio jet and an encroaching molecular gas can produce a
Fig. 3.—Schematic diagram of the jets and torus in NGC 1052. Double-sided jets are inclined by k57& with respect to the line of sight. The cross section of ageometrically thick molecular torus perpendicular to the jet is drawn. The inner face of the torus is ionized by irradiation from the central engine and causes free-freeabsorption. The X-ray dissociation region (XDR) is formed in the interlayer, where excited molecular gas performs maser amplification.
NARROW H2O MASER FEATURE IN NGC 1052 149No. 1, 2005
/ 18ALMA EA Science Workshp 2017 in Daejeon, Seiji Kameno : Molecular torus in the radio galaxy NGC 1052
Summary
18
• CND (radius~100 pc, Vrot ~ 150 km/s) • Absorption by accretion matter in molecular torus
- CO, HCN, HCO+, CS, SO, and CN- isotopologues : H13CN, HC15N- vib-excited HCN, HCO+
• Vertical structure of a geometrically thick torus- Molecular + XDR + plasma - Clumpy molecular clouds
ALMA + KVN is the best combination to quest mass accretion in radio galaxies
• Molecular emission/absorption line with ALMA • Locating absorbers with KVN
Discoveries from NGC 1052
We can estimate an approximate size of the foregroundabsorbing molecular gas using the relation =N f n LvH H2 2 , wherefv is the volume filling factor, nH2 is the molecular hydrogenvolume density, and L is the size of the molecular gas. Here wesimply assume fv = 0.01. Adopting NH2 of 10
24–1025 cm−2 fromour results and nH2=(0.1–2)×106 cm−3 in the Galactic CND(Smith & Wardle 2014), the resulting L is approximately on 1-pcscales. It is consistent with the approximate size of thecircumnuclear torus of NGC1052 (Kameno et al. 2001). Onthe other hand, the size of the innermost receding jet componentcan be estimated as 0.2–0.4 mas from the VLBA images at43GHz (Kadler et al. 2004; Sawada-Satoh et al. 2008), and itcorresponds to 0.02–0.04 pc in linear scale.
Based on our measurements of the radial velocities and thedistribution of HCN absorption features, the mass infall rate ofthe gas accretion toward the central engine can be estimated by
r= WM f R Vv in2
in˙ , where fv is the volume filling factor, Rin is aninfall radius, ρ is the mass density of the infalling materials, Vinis the infall velocity, and Ω is the solid angle of the torus fromthe center. Here we assume r = N m RH H in, where NH is thecolumn density of a hydrogen atom and mH is the mass of ahydrogen atom. Giving Rin = 1 pc, NH=1024–1025 cm−2, and
Vin = 200 km s−1 as the approximate velocity of the HCNabsorption feature with respect to Vsys ( -V Vp sys in Table 1),the derived mass infall rate ranges M=(47–470) fv(Ω/4π)M☉ yr−1. If (Ω/4π) takes a few 0.1 and fv is 0.01, it wouldbe comparable to the calculation of the accretion rate of10−1.39M☉ yr−1 from the hard X-ray luminosities by Wu &Cao (2006). We note that the X-ray luminosity indicates aninstantaneous accretion rate. However, our estimation of theaccretion rate in the molecular torus gives a long-term (Rin/Vin∼5000 year) activity. The coincidence of the two accretionrate values suggests the continuity of AGN activity inNGC1052.
5. SUMMARY
We have conducted 1 mas angular resolution observationstoward the HCN(1–0) absorption of the circumnuclar region inNGC1052 with the KVN. Two HCN absorption features areidentified and reach a depth of �10% from the continuum levelat 1656 and 1719 km s−1, redshifted with respect to Vsys. Wefind an NHCN of 1015–1016 cm−2 in the center of NGC1052,assuming an HCN covering factor of 1 and a Tex of 100–230 K.
Table 1Column Density of HCN (1–0) Absorption
Label Vp Vp − Vsys Δv ò tdv NHCN,100 NHCN,230(km s−1) (km s−1) (km s−1) (km s−1) (1014 cm−2) (1014 cm−2)
(1) (2) (3) (4) (5) (6) (7)
I 1656 149 31.7 9.3±0.6 9.5±0.6 50±3II 1719 212 52.9 19±1 20±1 101±5
Note. (1) Absorption feature ID, shown in Figure 1. (2) Peak velocity of the absorption feature. (3) Velocity with respect to Vsys. (4) Line width. (5) Velocity-integrated optical depth. (6) HCN column density with Tex=100 K. (7) HCN column density with Tex=230 K.
Figure 3. (a) Possible model of the oriented double-sided jet and the circumnuclear torus. The near side of the torus covers the receding jet side. (b) Schematic diagramfor the intersection of the circumnuclear torus in NGC1052. The torus has two-phase layers on the inner surface, the X-ray heated plasma region and the X-raydissociation region (XDR). HCN absorbing gas is located in the cooler molecule region next to XDR. HCN absorbing gas could be clumpy, and the line of sight passesthrough several clumpy clouds with a different radial velocity.
4
The Astrophysical Journal Letters, 830:L3 (5pp), 2016 October 10 Sawada-Satoh et al.