Cosmological growth of supermassive black holes: a synthesis model for accretion and feedback
Accretion in Protoplanetary Disks (conventional and...
Transcript of Accretion in Protoplanetary Disks (conventional and...
-
EC Daniel Perez-Becker
(Berkeley)
Accretionin Protoplanetary
Disks(conventional
and transitional)
-
pre-shock flow
heated interior
Calvet & Gullbring 98Muzerolle et al. 05Hartmann et al. 06
Herczeg & Hillenbrand 08
Pre-main-sequence stars accrete
Blue excess powered by accretion
-
Calvet et al. 05
Transitional Disks
τ10µm ∼ 20 − 200
τ10µm
= 0.05
TW Hyd
Calvet et al. 02
Brown et al. 08
LkHa 330
-
Holes are not empty
• Mild near-IR excesses in some sourcesτ10µm ∼ 0.002 − 0.05
• Many accreteṀ ∼ 0.1 × median T Tauri rate
• Inner molecular gas disksΣ(H2) > 0.1 g cm
−2 at ∼ 0.2AU
Najita et al. 07 Salyk et al. 07
-
Theories:
• Grain growth
• Clearing by companion
Puzzle:
Not mutually exclusive
1000 × smaller τ but comparable Ṁ
-
Clearing by companion (Transitional = Circumbinary)
Ireland & Kraus 08 (Keck AO)Artymowicz & Lubow 94
CoKu Tau/4Binary separationabinary ≈ 8 AU
≈ Hole radiusarim ≈ 10 AU
Ṁ∗ < 10−10M#/yr
D’Alessio et al. 05Najita et al. 07
Blake, Salyk, personal comm.
τ10µm < 0.002
No CO gas out to 2 AU
-
Mass can still accrete onto star(viscosity / pressure / eccentricity /
mass ratio)
Fast flow (up to radial free-fall)implies low surface density
Resolves the puzzle ofsimilar M but
1000x lower optical depth
.
Hanawa et al. 06
Clearing by companion (Transitional = Circumbinary)
-
Clearing by multiple planets
Simulation of 4-planet system in viscous disk
Right sign, too small magnitude‒ unless dust is also depleted
Zhu et al. 11 2D FARGO
-
Planets inside the hole do not explain disk accretion
Outer disk must stillbe viscous
-
Magneto-rotational instability (MRI)= linear instability
which drives turbulencein weakly magnetized,
outwardly shearing flows
1. Magnetic flux freezing (defeat Ohmic dissipation) (Fleming, Stone, & Hawley 00)
2. Good neutral-ion coupling (defeat ambipolar diffusion) (Blaes & Balbus 94; Hawley & Stone 98; Bai & Stone 11)
ReM ≡csh
η∝
nen
Requirements
> Re∗
M ≈ 102− 10
4
Am ≡ni〈σv〉in
Ω> Am
∗ ≈ 100
Disk accretion
Hawley 2000
1-100
-
Sideways
N, h
N, �
X− rays
h ≈ N/nH2Σ = Nµ
cosmic rays
1
Surface Layer Accretion by the MRI
Sources of ionization
1. Cosmic rays2. Stellar X-rays3. Stellar UV
Gammie 96
Glassgold et al. 97Sano et al. 00
Ilgner & Nelson 06 Bai & Goodman 09
Turner et al. 10Perez-Becker & EC 11aPerez-Becker & EC 11b
-
Galactic cosmic raysblocked by stellar wind
sunspot number
cosmic ray flux
interstellar
solar min
Svensmark 98Reedy 87
-
Pre-main sequence stars are X-ray luminous
Preibisch et al. 05Chandra COUP Orion survey
-
H+2 + H2 → H
+3 + H
H+3 + CO → HCO
++ H2
e−
dissociativerecombination
H + CO
chargetransfer
Mg+
Mg
βt ∼ 10−9
cm3s−1
e−
Mg
Mg
radiativerecombination
grain−
freemetal
Oppenheimer & Dalgarno 74Umebayashi & Nakano 83Glassgold et al. 86
βdiss ∼ 3 × 10−7
cm3s−1
βrec ∼ 4 × 10−12
cm3s−1
PAH-
neutralization onto charged particles
Mg
-
Polycyclic Aromatic Hydrocarbons (PAHs)
Allamandola et al. 99Geers et al. 06
PAH abundance ~0.01-10 ppb H2
(in diffuse ISM~1 ppm H)
-
10−32
10−26
10−20
10−14
10−8
10−2
αPA
H,e
or
αPA
H,M
+[c
m3s−
1]
10−5
10−4
10−3
10−2
10−1
1
Norm
alize
dch
arg
edis
trib
ution
−3 −2 −1 0 1 2 3Z
PAHs
Charg
edis
trib
ution
αPAH,eαPAH,M+
10−3
10−2
10−1
100
αgra
in,e
or
αgra
in,M
+[c
m3s−
1]
10−5
10−4
10−3
10−2
10−1
1
Norm
alize
dch
arg
edis
trib
ution
−40 −30 −20 −10 0Z
grains
Chargedistribution
α grain
,e
αgrain,M+
10−15
10−13
10−11
10−9
10−7
10−5
xe
or
(xM
++
xH
CO
+)
10−13 10−11 10−9 10−7 10−5
x�PAH
Possible PAHabundances
limxPA
H�x �PA
H xHCO +
limxPA
H�x �PA
H xe
limxPAH→0
xe, xHCO+
xM = 10−8
�PAH
xPAH
10−110−7 10−5 10−3 101
xM+ + xHCO+
xe
X-rays ⇒ HCO+, Mg+ egrain (1 μm)PAH (6 Å)
---
Charge Balance
-
0.01
0.1
1
10
100
Am
=(x
M++
xH
CO
+)β
inn
H2/Ω
0.01 0.1 1 10
Σ [g cm−2]
a = 3 AUxM = 10−8
Am∗
109 1010 1011 1012 1013nH2 [cm
−3]
1
102
104
106
108
1010
Re
=c s
h/D
Re∗
a = 3 AUxM = 10−8
0.01 0.1 1 10
Σ [g cm−2]
a = 30 AUxM = 10−8
Am∗
No PAH
Low PAH
High
PAH
Possible PAH
abundances
108 109 1010 1011nH2 [cm
−3]
Re∗
a = 30 AUxM = 10−8
Perez-Becker & EC 11asee also Mohanty, Ercolano, and Turner 11, in prep.
Ohm
ic d
issi
patio
nA
mbi
pola
r di
ffusi
on
region allowedby PAHs
Re
Am
0.01
0.1
1
10
100
Am
=(x
M++
xH
CO
+)β
inn
H2/Ω
0.01 0.1 1 10
Σ [g cm−2]
a = 3 AUxM = 10−8
Am∗
109 1010 1011 1012 1013nH2 [cm
−3]
1
102
104
106
108
1010R
e=
c sh
/D
Re∗
a = 3 AUxM = 10−8
0.01 0.1 1 10
Σ [g cm−2]
a = 30 AUxM = 10−8
Am∗
No PAH
Low PAH
High
PAH
Possible PAH
abundances
108 109 1010 1011nH2 [cm
−3]
Re∗
a = 30 AUxM = 10−8
(0.01-10 ppb H2)
X-ray ionization only
Field may be frozen to plasma ✔
But ions decoupled from neutrals ✗
-
= 0.1
= 0.01
= 10 3
= 10 4
MRI Prohibited
MRI Permitted
Am
<>
10 2 10 1 100 101 102 103 104100
101
102
103
104
10−11
10−10
10−9
10−8
10−7
10−6
Ṁ[M
⊙/yr]
1 10 100
a [AU]
Conventional Disk
Observed Rates
X-ray
α ≈1/(2β) and βmin(Am)
⇒ αmax (Am)
Bai & Stone 11
Kunz & Balbus 94 Hawley et al. 95
Desch 04Pessah 10
Bai & Stone 11
MRI with ambipolar diffusion
αmax (Am) and Am(Σ)
⇒ M.
X-rays + PAHs =weak accretion
Perez-Becker & EC 11ab
-
Far-Ultraviolet (FUV) Ionization
912 Å 1109 Å 1198 ÅH H2, C S
HST + FUSE(Bergin et al. 03; Herczeg et al. 02)
LFUV ~ 1030-1032
erg/s
-
FUV Ionization
912 Å 1109 Å 1198 ÅH H2, C S
Strömgren slab
}LFUVhν 4πa2 · ha ∼ nineαrec · hphotoionizations recombinations(cosmic abundance)Perez-Becker & EC 11ab
∼ 0.1
(
LFUV1030 erg/s
)1/2 (10−5
f
)
g/cm2
⇒ ΣMRI ∼ nH2µh
nH2
ni = ne = fnH2
-
1
10
102
103
104
Am
=(x
CII
+x
SII)β
inn
tot/
Ω
10−5 10−3 10−1
Σ [g cm−2]
Am∗
a = 3 AU
107 108 109 1010 1011ntot [cm−3]
1
102
104
106
108
1010
1012R
e=
c sh
/D
Re∗
Σ∗ Σ∗
LX = 1031 erg s−1
Re
Am
FUV ionization only
Field is frozen to plasma ✔Good ion-neutral coupling ✔
robust against PAHs(PAHs included at
maximum abundance)
sensitive to dust-to-gas ratio(vertical settling of dust)
Perez-Becker & EC 11b
Nτ=1 =1024 cm-2
[S/H] = -1
Nτ=1 =1022 cm-2
[S/H] = 0
-
102
103
104
105
χ
1 10 100
a [AU]
Nτ=1
VSG= 10
24 cm−2 � = 1/30
FUV
Wardle & Salmeron 11
Hall effect not important
-
FUV-ionizedsurface layerscan reproduceaccretion ratesat large radius
but not small radius
Perez-Becker & EC 11b
Ṁ ∼ 2 × 3πΣ∗ν
∼ 6πΣ∗αkT
µΩ
10−11
10−10
10−9
10−8
10−7
10−6Ṁ
[M⊙/yr]
1 10 100
a [AU]
ConventionalDisk
ObservedRates
FUV
max
βplas
ma<
1
X-ray
-
0.1
1
10
t rec
/t d
yn
1 10 100
a [AU]
FUV
Turbulent mixingof plasma to
greater depthscan extend
MRI-active layer
Ion recombinationtime vs.
dynamical time
Ilgner & Nelson 06b
-
10−12
10−11
10−10
10−9
10−8
10−7
10−6
Ṁtr
ans[M
⊙/yr]
1 10 100
arim [AU]
TransitionalDisk
FUV
X-ray
maxβ
plasm
a<
1
Murray-Clay & EC 07Kim et al. 09
Perez-Becker & EC 11abZhu et al. 11
FUV-driven MRI in Transitional Disks
Rim accretion ratereproduces
observations
Transport problemat small radii
could be solved bycompanions
-
Extra Slides
-
Is the required field super-equipartition?
Ṁ ∼ 〈BrBφ〉h/Ω
⇒ minB > 1G
(
Ṁ
10−8M"/yr
)1/2( r
AU
)−5/4
Possible for r > 1 AU
e.g., Bai & Goodman 09
β ≡PgasPmag
=8πnkT
B2< 1
(
Σ
0.1 g/ cm2
) (
10−8M"/yr
Ṁ
)
( r
1 AU
)
On the other hand the field cannot be too strong: β > 1
On the one hand, the field must be strong enough:
-
Inside-Out Accretion of Transitional Disks
-
• Dust thickness h/r ≤ R/r ̃ 10-2
• Inclination Ι ̃ 10-1• Warp Δ Ι / Ι < 10-1 (self-gravity) Δ Ι / Ι ̃ ‒10-1 (gas pressure)
Measuring dust layer thicknesses
(e.g., occulting circumbinary diskof KH 15D)
-
EC & Murray-Clay 07
Ṁ ∼12παN∗a2
rim(kT ∗)3/2
GM∗µ1/2
Rim controlsaccretion rate
X-ray driven MRI
@ 1 AU
• 10-100 x lower density than MMSN• Satisfies CO lower limits• Type II migration
slower than usual
For constant α,
But cannot explainorigin of hole
Σgas ∼ 10 − 100 g cm−2
Kim et al. 09Spitzer Cha I
-
Planet Clearing
Lubow & D’Angelo 06 Lubow et al. 99
Inner disk fills in
Ṁinner ≈ 0.1Ṁouter • But planetary migration reduces clearing efficiency• Short-lived solution (tdiffusion ~ r2 / ν ~ 104 orbits)• Perhaps multiple planets can help
Initialt = 700 orbits
neglecting migration
Initial profile set toanalytic steady state; note robustness of inner disk
-
But deeper correlations may exist ...
arim ∝ M2
∗
Ṁ∗ ∝ M2
∗
Ṁ∗ ∝ M2
∗
Same
holds fornon-transitional
disks
How to keep the inner hole
clear of dust?
Leaked dust might concentrate at , restoring :
Gapped(“pre-transitional”) disk
possible
a ! arim
τ10µm > 1
.
Kim et al. 09
-
1023 cm-2
tdiff ∼ a2rim / ν
ν = α csh
MRI simulations give 10-4-10-1
Mrim = 2πarim × 2h × N∗ µ
photo-ionization heating CO ro-vibrational cooling
230 K
*
Glassgold, Najita, & Igea 04EC & Murray-Clay 07
Ṁ ∼Mrimtdiff
∼
12παN∗a2rim(k T∗ )3/2
GM∗µ1/2
LXσXe−N
∗σX fheatn
4πa2rim
∼ ΛCO(T∗)
*
X-rays
}N*
T*a rim
X-rays
dustygasrim
MRI-active
Inside-Out MRI
-
tblow ∼1
Ω
(
1
Ωtstop
)
N*τrim ∼ 0.1
• Leaked dust might concentrate at , restoring : Gapped (“pre-transitional”) disk possible
a ! arim
τ10µm > 1
EC & Murray-Clay 07
-▽Ppoints in
Aerodynamic filter Radiation pressure
Gas is super-Keplerian
Dust is Keplerian ∴ dragged out
Keeping inner hole clear of dust
• Filtering is inefficient
Rice et al. 06
-
“Pre-Transitional” Gapped Disks
Espaillat et al. 08
Espaillat et al. 07
1600 K blackbody fit to excess
0.12 AU
< 0.15 AU
LkCa 15 46 AU
-
But deeper correlations may exist ...
˙
arim ∝ M2
∗
Ṁ∗ ∝ M2
∗
Ṁ∗ ∝ M2
∗
Why?
And does similarrelation hold
for debris disks?
Same
holds fornon-transitional
disks
-
(T̂ > 1)
Sustaining MRI at a
-
Outer diskphotoevaporationstarves inner disk
t=0
t=6 Myr
t=6.02 Myr
t=6.18 Myr
Alexander 08
-
x4e+
(
βgrβrec
)
x3e+
(
ζ
nβdiss
βtβrec
)
x2e−
(
ζ
nβdiss
βtβrec
) (
βgrβt
+ xmet,tot
)
xe−βt
βrec
(
ζ
nβdiss
)2
= 0
n = 2N/arim ζ =LXσXe
−NσX ξsecondary
4πEXa2rim
Estimating N*
xmet,tot = 10−6
10−7
10−8
EC & Murray-Clay 07
0
Am*
N*
-
Theories for transitional disks are not mutually exclusive
Planets + MRI +
smaller Ṁ∗ smaller τ10µmorigin
of viscosityMultiple planetsmight explainfactor of 10
and prolong Type II migration
Grain growth +Radiative /
aerodynamic blowout
smaller τ10µm
Imperfectclearing can
lead to gapped disks(e.g. LkCa 15)
-
0.01
0.1
1
10
100
Am
=(x
M++
xH
CO
+)β
inn
H2/Ω
0.01 0.1 1 10
Σ [g cm−2]
Am∗
109 1010 1011 1012 1013nH2 [cm
−3]
1
102
104
106
108
1010
Re
=c s
h/D
Re∗
LX = 1031 erg s−1
0.01 0.1 1 10
Σ [g cm−2]
Am∗
No PAH
Low PAHHigh
PAH
PossiblePA
H
abundances
109 1010 1011 1012 1013nH2 [cm
−3]
Re∗
Std. Model (a = 3 AU)
0.01 0.1 1 10
Σ [g cm−2]
Am∗
109 1010 1011 1012 1013nH2 [cm
−3]
Re∗
xM = 10−6
0.01
0.1
1
10
100
Am
=(x
M++
xH
CO
+)β
inn
H2/Ω
0.01 0.1 1 10
Σ [g cm−2]
Am∗
109 1010 1011 1012 1013nH2 [cm
−3]
1
102
104
106
108
1010
Re
=c s
h/D
Re∗
LX = 1031 erg s−1
0.01 0.1 1 10
Σ [g cm−2]
Am∗
No PAH
Low PAH
High
PAH
PossiblePA
H
abundances
109 1010 1011 1012 1013nH2 [cm
−3]
Re∗
Std. Model (a = 3 AU)
0.01 0.1 1 10
Σ [g cm−2]
Am∗
109 1010 1011 1012 1013nH2 [cm
−3]
Re∗
xM = 10−6
0.01
0.1
1
10
100
Am
=(x
M++
xH
CO
+)β
inn
H2/Ω
0.01 0.1 1 10
Σ [g cm−2]
Am∗
109 1010 1011 1012 1013nH2 [cm
−3]
1
102
104
106
108
1010
Re
=c s
h/D
Re∗
LX = 1031 erg s−1
0.01 0.1 1 10
Σ [g cm−2]
Am∗
No PAH
Low PAH
High
PAH
PossiblePA
H
abundances
109 1010 1011 1012 1013nH2 [cm
−3]
Re∗
Std. Model (a = 3 AU)
0.01 0.1 1 10
Σ [g cm−2]
Am∗
109 1010 1011 1012 1013nH2 [cm
−3]
Re∗
xM = 10−6
X-ray ionized MRI-active surface layer
Σactive ~ 1 g cm-2
Am ~ 1 (α ~ 10-3)
Perez-Becker & EC 10
At 3 AU,
-
0.01
0.1
1
10
100
Am
=(x
M++
xH
CO
+)β
inn
H2/Ω
0.01 0.1 1 10
Σ [g cm−2]
Am∗
108 109 1010 1011nH2 [cm
−3]
1
102
104
106
108
1010
Re
=c s
h/D
Re∗
xM = 0
0.01 0.1 1 10
Σ [g cm−2]
Am∗
No PAH
Low PAH
High
PAH
Possible PAH
abundances
108 109 1010 1011nH2 [cm
−3]
Re∗
Std. Model (a = 30 AU)
0.01 0.1 1 10
Σ [g cm−2]
Am∗
108 109 1010 1011nH2 [cm
−3]
Re∗
ζCR = 1/4 × 10−17 s−1
0.01
0.1
1
10
100
Am
=(x
M++
xH
CO
+)β
inn
H2/Ω
0.01 0.1 1 10
Σ [g cm−2]
Am∗
108 109 1010 1011nH2 [cm
−3]
1
102
104
106
108
1010
Re
=c s
h/D
Re∗
xM = 0
0.01 0.1 1 10
Σ [g cm−2]
Am∗
No PAH
Low PAH
High
PAH
Possible PAH
abundances
108 109 1010 1011nH2 [cm
−3]
Re∗
Std. Model (a = 30 AU)
0.01 0.1 1 10
Σ [g cm−2]
Am∗
108 109 1010 1011nH2 [cm
−3]
Re∗
ζCR = 1/4 × 10−17 s−1
Sideways
N, h
N, �
X− rays
h ≈ N/nH2Σ = Nµ
cosmic rays
1
Σactive ~ 0.1 g cm-2if no cosmic rays
At 30 AU,
Am ~ 1 (α ~ 10-3)
Σactive ~ 10 g cm-2if cosmic rays
X-ray ionized MRI-active surface layer
-
Far-UV (912-1100 Å) ionized MRI-active surface layer
H2 + hν→ H + HH + H + grain → H2 + grain
C + hν ⇄ C+ + e-
xCO = 10-4
Σactive ~ 0.01 g cm-2
At 3-30 AU,
Am > 102 (α ~ 0.1)
1
10
102
103
104
Am
=(x
C+)β
inn
H/Ω
0.01 0.1 1 10
Σ [g cm−2]
Am∗
109
1010
1011
1012
nH [cm−3]
1
102
104
106
108
1010
1012
Re
=c s
h/D
Re∗
fC = 10−4
a = 3 AULX = 1031 erg s−1
0.01 0.1 1 10
Σ [g cm−2]
Am∗
108
109
1010
1011
nH [cm−3]
Re∗
fC = 10−4
a = 30 AU
0.01 0.1 1 10
Σ [g cm−2]
Am∗
109
1010
1011
1012
nH [cm−3]
Re∗
xM = 10−6fC = 10−4
a = 3 AU
-
Planet Clearing
Run duration = 100 orbits ≪ Viscous time ~ 10000 orbits
Initial Σinner / Σouter = 0.01
∴ Hole in simulation reflects assumed initial conditions
Quillen et al. 2004
tdiff ∼ a2rim / ν
ν = α csh
0.004 (assumed)
tdiff
0.1MJ
10AU
arim =
-
3 AUX-ray3 AU
Far-UV30 AUX-ray+CR
30 AUFar-UV
αΣactive(g/cm2)T
(K)Ṁ
(M☉/yr)
10-118010-31
3010-30.1-10 10-11-10-9
0.1
0.1
0.01 300 4 x 10-11
0.01 300 10-9
-
200 AU
Protoplanetary Disks
disk mass ~ 0.001-0.1 stellar mass