Protoplanetary disksLeonardo Testiltesti@eso.org

• Origins of disks, overview and timescales• Disk structure and evolution• Observations: properties from continuum observations • Molecular gas properties from CO and isotopologues• Most likely not cover: Transition Disks and photo evaporation• Definitely not cover: dust evolution and planet formation(Warning: very selective material and focused mostly on submm)

mailto:ltesti@eso.org

Stars form in molecular clouds

molecular core

Molecular gas on the Galactic plane

Jeans criterion

M jeans ∝ T 3/2 n -1/2 → 2 Μsun for T=10 K, n=104 cm-3

As an order of magnitude

estimate, a star has

~one free-fall time

to reach its final mass

~ 105 y n5-1/2

From an observational perspective …

• Molecular cores are cold• T ~ 10 K

• and dense• n ~ 10 5 - 10 6 cm -3

The (once) elusive protostars

“an interstellar cloud undergoing inexorable gravitational contraction to form a single star”

—L. Spitzer (1948)

“Protostars are the Holy Grail of infrared astronomy”

—C.G. Wynn-Williams (1982)

Where is the disk coming from?Cores are rotating:

Centrifugal radius:(angular momentum conservation)

DensityStreamlines

angular momentum conservation r2Ω

Increase

s with

time fo

r

the dur

ation o

f the

collapse

phase

Rotation of coresMeasuring rotation

Goodman et al. (1993)

� = Erot

/Egrav

ALMAcaniden+fyacompactrota+ngcomponentaroundclass0protostar!

SMA

280AU

Ohashietal.2015

L1527Rota+onCurvemeasuredbyALMAOhashietal.2015

N

S

(UV excess)

Classical TTauri star

YSO SED classification

• Original classification I-II-II based on IR spectral index

• Class 0 based on FIR-submm• Many caveats, but still a useful

classification scheme

IR SED slope

Meyer+1997

Photospheres

Infrared excess

Extincted photospheres

Statistics and timescales

Evans et al. 2009

Statistics and timescales

Evans et al. 2009

Statistics and timescales

Evans et al. 2009

• In nearby star forming regions, we can estimate:• very roughly tII ~ 5-10 tI ~ 5-10 t0 or ~106 yr : ~105 yr : ~104 yr

From Cores to Planetary Systems

LCore

Disk

Debris Disk

From Cores to Planetary Systems

Core

Disk

Debris Disk

Inner disk clearing:e-folding time t~2-3 Myr

(Her

nand

ez e

t al.

2007

)

Our own Solar System

• CAI• Formed early

• Condrules• After 1-2 Myr• Spread ~3Myr

• Matrix• Sub-um particles• Cement the body

Our own Solar SystemThe accepted view has recently been challenged by recent measurements of isotopic abundances in CAIs and Condrules

(Con

nelly

et a

l. 20

12)

• CAI• Formed early

• Condrules• After 1-2 Myr• Spread ~3Myr

• Matrix• Sub-um particles• Cement the body

Protoplanetary disks

100 AU

Vertical structure

• Hydrostatic equilibrium: pressure vs gravity

€

ddz

ρ(z)kT(z)µmp

#

$ % %

&

' ( ( = −ρ(z)ΩK

2 z

€

kTµmp

dρ(z)dz

= −ρ(z)ΩK2 z

€

ρ(z) = ρ0 exp −z2

2h2$

% &

'

( )

€

h = kTr3

µmpGM*

• Using an isothermal approximation, we can find a gaussian solution for the density

• h

viscosity and radial structure• The vertical timescale (h/cs) is much shorter than the

radial evolution timescale (can decouple the two processes)

• Assumption: vr

• Alpha-disk after Shakura & Sunyaev (1973)viscosity and radial structure

• Self similar solution after Linden-Bell & Pringle (1973) • if T~r-q then n~r(3-2q)/2 (assuming alpha~const)

€

ν =αcsh

€

α = 0.001...0.1

€

ν =αcs2

ΩK

Viscous disc similarity solution

�(R, t) =Md(0)(2� �)

2⇥R20r�

⇤�(5/2��)

2�� exp

✓�r

2��

⇤

◆

Lynden-Bell & Pringle (1974)

t⌫ =R20

3(2� �)2⇥0r = R/R0 � = t/t⌫ + 1

Ṁacc =Md(0)

2(2� �)t�⇥

�(5/2��)2��

� / R�

Viscous disc similarity solution

�(R, t) =Md(0)(2� �)

2⇥R20r�

⇤�(5/2��)

2�� exp

✓�r

2��

⇤

◆

Lynden-Bell & Pringle (1974)

t⌫ =R20

3(2� �)2⇥0r = R/R0 � = t/t⌫ + 1

Ṁacc =Md(0)

2(2� �)t�⇥

�(5/2��)2��

� / R�

Viscous disc similarity solution

�(R, t) =Md(0)(2� �)

2⇥R20r�

⇤�(5/2��)

2�� exp

✓�r

2��

⇤

◆

Lynden-Bell & Pringle (1974)

t⌫ =R20

3(2� �)2⇥0r = R/R0 � = t/t⌫ + 1

Ṁacc =Md(0)

2(2� �)t�⇥

�(5/2��)2��

� / R�

Viscous disc similarity solution

�(R, t) =Md(0)(2� �)

2⇥R20r�

⇤�(5/2��)

2�� exp

✓�r

2��

⇤

◆

Lynden-Bell & Pringle (1974)

t⌫ =R20

3(2� �)2⇥0r = R/R0 � = t/t⌫ + 1

Ṁacc =Md(0)

2(2� �)t�⇥

�(5/2��)2��

� / R�

• evolution of a thin annulus (gamma=1.0)viscosity and radial structure

Log(R)

Log(

Sigm

a)

• full disk evolution (gamma=1.0)viscosity and radial structure

Log(R)

Log(

Sigm

a)

• Data generally consistent with viscous evolution and Macc

Mass accretion rate

A more realistic picture

• Data generally consistent with viscous evolution and Macc

• Very slow!

Viscous evolution

Disk photoevaporation

• Energetic stellar radiation heats the disk surface• Unbound material flows away from the system

(Arm

itage

)

Photoevaporation cartoonSimple Theory of Photoevaporation from internal EUV radiation1. EUV radiation ionises and heats the disc atmosphere: bound atmosphere at r < rg; mass loss in thermal wind at r > rg. Ioniaing flux dominated by scattered stellar photons from the wind atmosphere (diffuse field)

2. Accretion rates drop below the mass-loss rates and a gap opens in the disc close to rg. The outer and the inner disc become decoupled. The inner disc is starved.

3.The inner disc drains rapidly on to the star an inner hole (transition disc is produced and the direct EUV flux from the star quickly photoevaporates the outer region.

Monday, 7 May 2012

• Consistent with disk evolution timescales• (… need to add planet formation …)

Evolution of Mdisk

Observations

Spectral Energy Distributions (SEDs)The SED shows the energy emitted per logarithmic wavelength

interval. Plotting just the flux may be misleading:

Energy:

€

Fν dν = Fν Δν

erg cm-2 Hz-1

Spectral Energy Distribution

• Balance energy losses Q- and source Q+:viscous heating

Q-=2sTeff4 (two sides, each radiate as a BlackBody)

Q+=(3/4p)Macc(WK)2 (viscous heating)Q-=Q+

Teff4=3/(8sp)Macc(WK)2Teff~r-3/4

Tmid4=tRossTeff4 tRoss=SkRoss/2

Wien

Intermediate region

Rayleigh-Jeans

τν∝Σ(r)κν Σ(r)∝r-p κν∝κoνβ

SED for a locally isothermal disk

λ

νFν

ν(4q-2)/qνFν

λ

τν∝Σ(r)κν Σ(r)∝r-p κν∝κoνβ

SED for a locally isothermal disk

ν3+β

τν∝Σ(r)κν Σ(r)∝r-p κν∝κoνβ

SED for a locally isothermal disk

Beckwith et al. (1991)

Viscous heating provides a poor fit of protoplanetary disc temperature:real disks are warmer than expected in the outer regions!

τν∝Σ(r)κν Σ(r)∝r-p κν∝κoνβ

“flat” passive diskIrradiation flux:

€

Firr =αL*4π r2

Coincidentally, same profile: no good!

The flaring angle:

€

α ≅0.4 r*r

€

T = 0.4 r*L*4πσ r3

$

% &

'

( )

1/ 4

€

T ∝ r−3 / 4

τν∝Σ(r)κν Σ(r)∝r-p κν∝κoνβ

“flared” passive diskIrradiation flux:

€

Firr =αL*4π r2

Can work……depending on hs(r)

The flaring angle:

€

α = r ∂∂r

hsr

$

% &

'

( ) →ξ

hsr

€

T 4 = ξσhsL*4π r3

€

hs = χ h

Flared disks: detailed models

Global disk model...

... consists of vertical slices, each forming a 1D problem. All slices are independent from each other.

[K. Dullemond]

Including viscous heating

Including viscous heating

(Macc~10-8 Msun/yr @ 1 Myr)

“flared” disk emission

Flared disks: which observations probe what?

Mid-IR imaging

Scattered light

IR SpectroscopyPAH

Emission

[C. Dominic]

Submm/radio: Entire Disk

Disk masses: (sub)mm continuum• MD ~ 0.01-0.1 Msun• MD/Mstar ~ 0.03

• F1mm ~ Bn(T) k1mm MD

Disk masses - Taurus

• Full Taurus survey, pre-ALMA: low detection rate M*

Disk masses: ALMA

Disk masses: ALMA

LTesti | ALMA Development | STC Apr 23, 2017

Forming the TRAPPIST-1 system

Young Brown Dwarf disks ➢Where is the material to form

TRAPPIST-1? ➢Planet formation much faster

than we think or TRAPPIST-1 unreasonably rare…

Examples of pre-ALMA results

• Data generally well described, note limited angular resolution

Isella+2009

Examples of pre-ALMA results

Isella+2009 Williams&Cieza 2011 Andrews+2010

• Extract N random samples from “Taurus” applying the same selection biases as expected in the other region. Method first applied by Andrew+2013

Examples of pre-ALMA results

Isella+2009 Williams&Cieza 2011 Andrews+2010

• Extract N random samples from “Taurus” applying the same selection biases as expected in the other region. Method first applied by Andrew+2013

Accretion

• Direct measurement of accretion: energy released in the collision

• Indirect measurement: emission lines from accretion columns

(UV excess)

L. Hartmann

Manara+2014

Gullbring+2000

Accretion

• Lacc measured directly from the UV excess luminosity (with a correction factor for FUV/EUV

(UV excess)

L. Hartmann

Gullbring 1998

Rm ~ 5 Rstar

Rigliaco 2011

ALMA - Md

XSho

oter

- M

acc

Man

ara+

2016

ALMA - Md

XSho

oter

- M

acc

Mul

ders

+201

7

Accretion-Md connection

• Almost, but not quite…

Dul

lem

ond,

Nat

ta, T

esti

2006

30 au

100 =1

3600degree

1 au = Sun-Earth distance

orbit of neptune = 30 au

gaps

thermal emission of solid particles ≲ 1 cm Kwon et al. 2011

thermal emission of solid particles ≲ 1 cm ALMA Partnership 2015

Disks with ALMA

• Not as smooth and simple as advertised so far: dust evolution and dust-gas-planets interaction rule (another KES, sometime)

Summary• Disk formation and initial properties

• Process is understood in broad terms• Details are tricky: gas and dust initial content and

properties? Amount of L-loss during collapse?

• Disk properties• angular momentum loss drives early evolution:

viscosity (origin? properties?) and/or winds?

• photoevaporation and/or planet formation drives disk dissipation

• Observing dust is easy, gas a bit more tricky, understanding what observations tell us requires models!