Starbursts & Super Star Clusters J. Gallagher U. Wisconsin-Madison with L. J. Smith, M. Westmoquette (UCL), R. O’Connell (UVa), R. de Grijs (U. Sheffield)

Starbursts&

Super Star Clusters

J. Gallagher

U. Wisconsin-Madison

with L. J. Smith, M. Westmoquette (UCL), R. O’Connell (UVa), R. de

Grijs (U. Sheffield)

MODEST-6 Workshop- 29 August 2005

Super Star Clusters (SSCs):Upper Mass Range for Young Compact

Clusters--Densest Single Generation Stellar Systems--Endpoint of Star Formation

Dissipational Sequence • Stellar mass >104 Msun

• Half light radius R1/2 ≤ 5-7 pc

• Within R1/2 N*(1 Msun) ~104 - 106 pc-3

• Ages << globular clusters 0-few 100 Myr

• Common in starbursts--sometimes tightly packed in starburst “clumps”

0.01 0.1 1

Starburst Scale (kpc)

Inte

nsit

y

low

high

NGC1569

M82M83

Possible Starburst Scale-Intensity

ULIRGs

OB Assoc

Spiral arm

Orion

1012 L

1010 L

108 L

N1275Starbursts

Main SFR density starburst contribution from moderate mass galactic starbursts:

“downsizing”.

Starbursts are not simply scaled-up versions of normal galactic disks.

Measuring Parameters

• Linear size R--high angular resolution, 0.1 arcsec or better needed for D<10 Mpc; half light radius.

• Age tcl--spectral energy distributions; colors okay if broad wavelength coverage; NIR alone difficult

• Luminosity Lcl()--photometry & extinction correction; zone of radiative influence for ionization and mechanical luminosities

• Chemical abundances Z*--stellar or HII spectra• Dynamical mass--stellar velocity dispersion,

requires cool stars (tssc > 6-7 Myr)• Stellar mass function--resolved cluster or Mdyn/L

with age & model from measured Lcl(, Z*, tcl, Mcl*)

Archetype: 30 Doradus:

small super star cluster or

“SSC”

Berstein & Novaki 1999, APOD

R136 R

ad

ial P

rofi

le

€

ρ(R) ≈ρ 0 R−3

Hunter et al. 1995ApJ, 448, 179

A

B

10

NGC 1569: Dwarf Starburst GalaxyP. Anders, U. Goettingen; data HST: ESA/NASA

De Marchi et al. 1997, ApJ, 479, L27

A = Double Cluster

WR*

Composite spectra:Mixture of ages--high

mass stars, >30-40 Msun

present in SSCs;RSGs in optical of

cluster A

Ho & Filippenko 1996, ApJ, 466, L83

Cool star

Cluster A

Smith & Gallagher 2001, MNRAS, 326, 1027

NGC 1569 - NIR with WIYN 3.5-m Telescope; Natural Seeing: SF Patterns: SSCs Embedded in Young Star Clouds

B

A

SSCs10

NGC 1569--3 SSCs in ~10 Myr--Feedback induced shift in SF

mode?Increased pressure of SSC A a dominant

factor?SFR declines as dense ISM exhausted-

and ejected? GMC formation vs. destruction?

Is the formation of SSCs a statistical process as the

stellar IMF appears to be:More clusters= higher upper mass limit?(Yes in Antennae--but are these compact

SSCs?(M82 ??)

ORIs SSC formation in some cases a result of

a feedback enhanced mode of star formation?(NGC1569?)

NGC1569 SSCs NOT detected as luminous X-ray sources: L(x) ≤ 2 x 1036 erg/s

Martin, Kobulnicky, Heckman, 2002, ApJ, 574,663

M. W

estm

oque

tte

(UC

L),

J. G

alla

gher

(U

W),

L. J

. Sm

ith

(UC

L)

Wit

h N

ASA

/ESA

and

WIY

N O

bse

rvat

ory/

NSF

M82: HST WFPC2 + WIYN

M82-nearby giant starburst

Stellar disk

J. Gallagher & L. J. Smith

M82F

M82: 3.5-m WIYN Telescope I-band

M82 view from the ground:

A VERY disturbed galaxy--

bright starburst

clumps: dust and

superimposed SSC “stars”

D=3.6 MPC 1 arcsec

≈19 pc

M82- clump A

M82-SSCs F &

L

M82- clump B

M82 SSCs & Starburst Clumps: V-band WFPC2

M82-A1 SSC: 106 M - r1/2~2-3 pc - t<10 Myr>> L. Smith Talk for Details! <<

Star

burs

t Clu

mp

A

~10 M

yr

Melo

et al

. 200

5, ApJ

NGC 7673 starburst-WFPC2

Homeier, Gallagher, Pasquali

Clumping of compact young star clusters-a step beyond super OB associations?

Characteristic of unstable gas-rich disks subject to intense star formation?

D=40 Mpc; MV=-20

Background galaxy

~8 kpc

Hubble Ultradeep Field: High-z clumpy & compact starbursts--a key early star formation mode

Chandra X-ray contours: wind mass loading + thermalization

Basic astronomy: Astrometry key to IDs,especially as distance increases…

Chandra vs clusters in Antennae: Fabbiano et al. 2002, ApJ, 577, 710

1 arcsec= 73(D/15 Mpc) pc

M82-Chandra X-ray vs HST NIC IR: X-Ray Shocked Winds & Binaries in SSCs?

Kaa

ret

et a

l. 20

04 M

NR

AS,

348

, L28 1 arcsec

Most SSCs not luminous X-ray

sources.

Wind thermalization low within clusters (clumpy winds???)

BH binaries ejected

Implications for intermediate mass BH

growth--nuclei vs. “field”

Age = 60 20 Myr W

HT

sp

ectr

osco

py:

Gal

lagh

er &

Sm

ith

199

9, M

NR

AS

, 304

, 540

25 a

rcsec

HST angular resolution essential to measure sizes!

430 p

cM82-F

€

M = 7.5σ 2r(half mass)

Gr(half mass) = 3.3 pc

σ =13.4 km /s

M =1.2 ×106 Msun

Sm

ith

& G

alla

gher

200

1, M

NR

AS

, 326

, 102

7M82-F: WHT Echelle Spectra & Mass: A

Doomed SSC?SSCs have the mass and size of globular clusters.

Star formation at the high density extreme of the interstellar gas cloud dissipation sequence.

Roles of initial conditions & environment in survival? SFE

vs Mcl?

Smit

h &

Gal

lagh

er 2

001,

MN

RA

S, 3

26, 1

027 L/M vs age of super star

clusters Appears to lack low mass stars!

BUTmass

segregation?

32

1

0.1

M_min

M82-F

?

McCrady, Graham, & Vacca 2005, ApJ, 621, 278

“Using PSF-fitting photometry, we derive the cluster’s light-to-mass (L/M) ratio in both near-IR and optical light and compare to population-synthesis models.The ratios are inconsistent with a normal stellar initial mass function for the adopted age of 40–60Myr, suggesting a deficiency of low-mass stars within the volume sampled.King model light profile fits to new Hubble Space Telescope ACS images ofM82-F, in combination with fits to archival near-IR images, indicate mass segregation in the cluster. As a result, the virial mass represents a lower limit on the mass of the cluster.”

?

5 pc

lossStar

Tides

GalacticIMF

Age

Massdensity

ISM

Gallagher & Grebel 2003, IAU 217

ComplexSta

r Cluste

r Ecology:

AffectsSurvivalRates!

Low M/L in an older SSC--signature of impending disruption? Why do GCs survive when mortality

rates seem high in nearby SSC systems?

Extra binding (DM???)--Special Formation Conditions??

Chandar et al. 2005, ApJ, 628, 210

Starburst field FUV spectra systematically “older” than SSCs--> substantial cluster dissolution within ~10 Myr?

de Grijs, O’Connell, Gallagher (2001)

BUT Many M82 region B star clusters have colors consistent with ages of near 1 gyr. (also

Parmentier, de Grijs & Gilmore 2003, MNRAS, 342, 208) ->> Evidence of long lived clusters

and multiple bursts associated with orbital period of M82.

M82 Starburst Clump B~1000 Myr clusters

But where are the 2000 Myr clusters?

Cluster Dynamical Evolution: Mass Segregation

€

tmass ∝ trhm * /mu* ≤ 0.1trh

≈70(M /105)1/ 2 MyrCompact young star clusters may be unstable

againstmass segregation. Primordial mass segregationpotentially amplified. IMF & survival complex

relationship.

M82-F as example. Appears likely candidate for disruption. Did low mass stars form in a more extended region?

Summary

• SSCs are commonly produced in intense star forming events with masses up to and beyond 106 Msun and R1/2 ~ 2-5 pc.

• Densest stellar cluster--an extreme of star formation.• SSCs cluster to make “starburst clumps” where SSC-SSC

interactions are possible and which drive galactic winds. • SSCs contain a full range of intermediate-high mass stars.• Central stellar densities can exceed 105 stars/pc3.

• Mass segregation may have a major influence on the observed properties (low M/L ratios) and evolution (dissolution) of SSCs. PDMFs thus are uncertain.

• Statistics of SSCs suggest high early disruption rates in starbursts, although in some cases significant numbers of clusters reach ~1 Gyr in age.

• SSCs do not appear to generally host strong X-ray sources, suggesting that SNe II binaries are ejected.

M82--11.7 microns--compact star forming regions

Lipscy & Plavchan 2004, ApJ, 603, 82

NGC1569-photo ages: Anders et al. 2004, MNRAS, 347, 17

2002 A&A, 381, 825

Merl

in:

com

pact

rad

io s

ou

rces

SNRs in NGC1569: Triggered SF?

McCrady, Gilbert, & Graham 2003, ApJ, 596, 240

Cluster Dynamical Evolution: Crossing Time

€

tcross = 2rh /(σ 3) ~1 Myr

∝ 1/ρ

Cluster crossing times typically 1% or less of galactic orbital periods and 10% or less of massive star evolutionary time scales.

We are dealing with systems that experience

substantial dynamical evolution: SSCs not

durable?

Clusters & Dynamical Evolution: Two Body Relaxation

€

trh ≈ 700(M /105)1/ 2(rrh /5 pc)3 / 2 m*−1C Myr

C = (lnΛ/10)−1 ≈1

Time scale for two-body relaxation to become important--a fundamental reference time in a star cluster:

But time scale varies with position and mass:

€

tr0 ∝σ 3n*−1m*

−2

Cluster cores evolve rapidly, especially if massive stars Preferentially form near the cluster center.

O. Gerhard 2000, Massive Stellar Clusters, p12

Martin, Kobulnicky, Heckman 2002, ApJ, 574, 663

V(HII) ~ 100 km/s--Slow optical wind

X-ray sources & clusters--diffuse hot gas backgrounds

M82 - SUPERWIND!

H +[NII]HST/WFPC2

A

C

Smith & Gallagher 2001, MNRAS, 326, 1027

Dwarf starbursts: NGC1705

Nuclear starbursts: M83

Interactions: M82

Major mergers: NGC6240

Tosi et al WFPC2

Harris et al WFPC2

Pasquali et al. WFPC2

Gallagher et al. WIYN

M82: Radio (Rodriguiz-Rico et al. 2004, ApJ, 616, 783) vs. HST F160W NICMOS Image OR

Perils of 1 arcsec rms positions

+ SNR x HII region