1) Are disks predicted? Theories of HM SF

2) Are disks observed? Search methods

3) Observational evidence disks VS toroids

4) Open questions and the future: ALMA, etc.

Search for Disks around YoungHigh-Mass Stars

Riccardo CesaroniINAF-Osservatorio Astrofisico di Arcetri

Existence of disks: Theory

Disks are natural outcome of infall + angular momentum conservation, however:

• B field magnetic braking, pseudo disks?• Ionization by OB stars photoevaporation?• Tidal interaction with cluster truncation?• Merging of low-mass stars destruction?

Disks in OB protostars might not exist!

Good news: all theories predict circumstellar disks! Different models of high-mass star formation (core accretion, competitive accretion, …), but all predict circumstellar disks of ~100-1000 AU

See e.g. Bonnell 2005, Krumholz et al. 2007, Keto 2007, Kuiper et al. 2010

stars up to 137 MO through disk accretion

1 pc clump collapsecompetitive accretion

Bonnell (2005)

Zoom in

time

core accretionin 0.2 pc clump

Krumholz et al. (2007)

disk

50 AU

molecular gas

ionized gas

density & velocity of gas around O9 star (Keto 2007)

Bad news: all theories predict circumstellar disks!

Disks existence not sufficient to choose SF theory

Disks may keep memory of formation process disks properties needed to discriminate between SF models

< 100 AU resolution necessary, i.e. < 0.1” ALMA, EVLA, eMERLIN, VLBI, VLTI, …

The search for disks

Many searches in the last decade need targets & toolsSelection criteria for targets:• Bolometric (IRAS) luminosity > several 103 LO

high-mass (proto)star• Association with outflow likely disk?• Presence of massive (> 10 MO), compact (< 0.1 pc)

molecular core deeply embedded (young) high-mass object

• In some cases maser and/or UCHII OB stars

Tools adopted:• Thermal lines of rare (low-abundance) molecules trace high-

density, high-temperature gas in disk• H2O, CH3OH, OH, SiO maser lines mas resolution• (sub)mm continuum disk mass• IR continuum/lines disk emission and absorption• cm continuum and RRL ionised accretion flow

Diagnostic:• Flattened (sub)mm core perpendicular to jet/outflow• Velocity gradient perpendicular to associated outflow• Peculiar (Keplerian) pattern in position-velocity plot• Dark silouhette in near-IR against bright background• Elongated emission in the mid-IR perpendicular to bipolar

reflection nebula

CepA HW2Jimenez-Serra et al. (2007,2009)

PV plot along disk

Keplerian rotation about 18 MO star

thermal jet

disk

B field (23 mG)CH3OH masers

Vlemmings et al.(2010)

SO2

600 AU

1000 AU

rel. D

ec.

[m

as]

NGC7538 IRS1 N Pestalozzi et al. (2004, 2009)

model of Keplerian disk

around 31 MO star

CH3OH maser

PV plotalong disk

diskplane

M17Chini et al. (2004)

Nuerberger et al. (2007)

2.2 µmcontinuum

H2 jetdisk

2 µm lines

4.5µmemission

disk

J23056+6016Quanz et al. (2010)

H K’bipolarnebula

12COblue outflow lobe

red & blueC18Odisk

Major problems:• Velocity gradient may be expansion instead of rotation• Outflow multiplicity and/or precession• Masers sample only few lines of sight• (sub)mm & IR continuum: no kinematical info• IR lines: so far limited spectral resolution

Possible solutions:• High angular & spectral resolution accurate PV plots

Keplerian rotation (close enough to star)• Maser proper motions 3D velocity

Combine as many tools as possible!

IRAS 20126+4104Cesaroni et al.Hofner et al.

Sridharan et al.Moscadelli et al.

Image: 2µm cont.

--- OH maser

H2O masers

1000 AU

Keplerian rotation+infall:

M*=10 MOMoscadelli et al. (2010)

CH3OH H2O200 AU

jet

disk+jetdisk

Distance measurement to IRAS 20126+4104 withH2O maser parallax (Moscadelli et al. 2010)

d = 1.64±0.05 kpc

Observational results

• Evidence for rotation/flattening in ~42 molecular cores:~26 disks Keplerian rotation in ~10 of these

~16 rotating toroids velocity gradient perpendicular to outflow/jet, but not Keplerian

PV plots of candidateKeplerian disks

in high-mass stars

13CO

-0.5”0.5”1” 0NH3(1,1)

CH3OH

IRAS23151

NGC7538

IRAS20126

CepAHW2CH3OH

NGC7538S

DCN

C17O

AFGL5142

AFGL490

IRAS18566

M17

W33A

CO v=2-0

disk

disk

disk

model 2.2µm VLT

diskmodel2.2µm VLT

model 2.2µm VLTI

disk

IR detected disks

AFGL2591

2.1µm speckle

2.2µm Subaru

J23056 IRAS20126

M17UC1

M17

19µm SubaruHD200775

IRAS13481

2.2µm UKIRT

Kraus+2010

Quanz+2010

Sridharan+2005

Steinecker+2006

Nielbock+2007

Kraus+2010

Okamoto+2009

Velocity fields of rotating toroids

C

G24 A1

G24 A2

G31.41

G19.61G10.62

G327

G351

G305

G28.20

CH3CN

CH3CN

CH3CNNH3 NH3

Toroids• M > 100 MO

• R ~ 10000 AU• L > 105 LO O (proto)stars• small tacc/trot

non-equilibrium, circum-cluster structures

Disks• M < a few 10 MO

• R ~ 1000 AU• L ~ 104 LO B (proto)stars• large tacc/trot

equilibrium, circumstellar structures

disks

toroidsBeltran et al. (2010)

Open questions

• When do disks appear? 1 disk/toroid in IR-dark cloud• Role of magnetic field? 2 toroids with B parallel to

rotation axis B may play crucial role• Why no (Keplerian) disks seen in O stars?

Ionized by OB stars? Unlikely: too slow and rotation in ionized gas detected in G10.62 (Keto & Wood 2005)

Truncated by tidal interactions in cluster? Maybe, but numerical simulations needed

Too far? ALMA and EVLA should tell us!Too deeply embedded in toroids? Optically thin (low

abundance i.e. high density) tracers needed, but line forest may fool even ALMA! VLBI of masers may help

IRDC18223-3Fallscheer et al. (2009)

CH3OH

5000 AU

disk model

small-scalevelocity field

large-scale bipolar outflow

disk-modelvelocity field

0.2 pc

Open questions

• When do disks appear? 1 disk/toroid in IR-dark cloud• Role of magnetic field? 2 toroids with B parallel to

rotation axis B may play crucial role• Why no (Keplerian) disks seen in O stars?

Ionized by OB stars? Unlikely: too slow and rotation in ionized gas detected in G10.62 (Keto & Wood 2005)

Truncated by tidal interactions in cluster? Maybe, but numerical simulations needed

Too far? ALMA and EVLA should tell us!Too deeply embedded in toroids? Optically thin (low

abundance i.e. high density) tracers needed, but line forest may fool even ALMA! VLBI of masers may help

G31.41+0.31

Cesaroni et al.in prep.

W51e2

Tang et al.(2009)

Keto&

Klaassen(2008)

Girart et al.(2009)

Open questions

• When do disks appear? 1 disk/toroid in IR-dark cloud• Role of magnetic field? 2 toroids with B parallel to

rotation axis B may play crucial role• Why no (Keplerian) disks seen in O stars?

Ionized by OB stars? Unlikely: too slow and rotation in ionized gas detected in G10.62 (Keto & Wood 2005)

Truncated by tidal interactions in cluster? Maybe, but numerical simulations needed

Too far? ALMA and EVLA should tell us!Too deeply embedded in toroids? Optically thin (low

abundance i.e. high density) tracers needed, but line forest may fool even ALMA! VLBI of masers may help

tidal destructionrotational period

photo-evaporation

Cesaroni et al. (2007)

Open questions

• When do disks appear? 1 disk/toroid in IR-dark cloud• Role of magnetic field? 2 toroids with B parallel to

rotation axis B may play crucial role• Why no (Keplerian) disks seen in O stars?

Ionized by OB stars? Unlikely: too slow and rotation in ionized gas detected in G10.62 (Keto & Wood 2005)

Truncated by tidal interactions in cluster? Maybe, but numerical simulations needed

Too far? ALMA and EVLA should tell us!Too deeply embedded in toroids? Optically thin (low

abundance i.e. high density) tracers needed, but line forest may fool even ALMA! VLBI of masers may help

Assumptions:HPBW = Rdisk/4

FWHMline = Vrot(Rdisk)

Mdisk Mstar

same <Ncol> in all disks

TB > 20 K

obs. freq. = 230 GHz

5 hours ON-source

spec. res. = 0.2 km/s

S/N = 20

edge

-on

i = 35

°

circumstellardisks

Keplerian

Assumptions:HPBW = Rdisk/4

FWHMline = Vrot(Rdisk)

Mdisk Mstar

same <Ncol> in all disks

TB > 20 K

obs. freq. = 230 GHz

5 hours ON-source

spec. res. = 0.2 km/s

S/N = 20

no s

tars

edge

-on

i = 35

°

Simulations of disksaround 8 MO star

Krumholz et al. (2007)

NH3 with EVLA

CH3CN(12-11) with ALMA

cont.+

line

cont.subtr.

Open questions

• When do disks appear? 1 disk/toroid in IR-dark cloud• Role of magnetic field? 2 toroids with B parallel to

rotation axis B may play crucial role• Why no (Keplerian) disks seen in O stars?

Ionized by OB stars? Unlikely: too slow and rotation in ionized gas detected in G10.62 (Keto & Wood 2005)

Truncated by tidal interactions in cluster? Maybe, but numerical simulations needed

Too far? ALMA and EVLA should tell us!Too deeply embedded in toroids? Optically thin (low

abundance i.e. high density) tracers needed, but line forest may fool even ALMA! VLBI of masers may help

Furuya et al. (2008)

CH3CN

Sanna et al. (2010)

rotating toroid deeply embedded disk?

CH3OH masers

1.3cm cont.