The Obscured Growth Phase of Black
Holes in Distant Massive Galaxies
David M Alexander (Durham)
What I’m not Going to Talk About:
Robust Identification of z~2-2.5 Compton-thick Quasars
Optical AGN
IR AGN
Compton-thick quasars (LX>1044 erg/s)
at z~2 are as numerous as unobscured
quasars: extending Daddi et al. (2007)
to confirming individual C-thick AGNs
What I will Talk About:
Black-hole-Galaxy Growth in z~2 Starbursts and Quasars
“Weighing the Black Holes
of z~2 Submillimeter
Galaxies and Exploring their
Evolutionary Status”
D.M. Alexander et al. AJ,
submitted
“Is there an Evolutionary
Link between Quasars and
Submillimeter Galaxies?”
K. Coppin et al. in prep.
Today’s most massive galaxies hint of a violent past
Heavens et al. (2004)
Formation must have been
M87 distant, rapid, and luminous
Also need to grow a massive black hole
Probe black-hole growth
with AGN activity
All massive galaxies appear to host a massive black hole
=> all galaxies have undergone luminous AGN activity in the past
Black Hole-Stellar Growth
Cannot “age” a black
Action: AGN activity
MBH = 0.15 % Mbulge hole, as you can age
stars but the tightness
of the black-hole-
spheroid mass
relationship suggests
they may have grown
concordantly
Challenging tests for
Tremaine et al. (2002) structure-formation
models
Action: Star formation
Submillimeter/Millimeter: efficient selection of the
most bolometrically luminous far-IR galaxies in Universe
850 micron SCUBA image
(Blain & Longair 1996)
Lots of
them!
Before SCUBA2,
submm will miss hot
Coppin et al. (2006) ultraluminous sources
Submillimeter/Millimeter: efficient selection of the
most bolometrically luminous far-IR galaxies in Universe
850 micron SCUBA image
(Blain & Longair 1996)
Hughes et al. (1998) Lots of
them!
Before SCUBA2,
submm will miss hot
Smail et al. (2002) ultraluminous sources
Galactic Properties…
Chapman et al. (2003, 2005) Strongly clustered
Distant: typically Swinbank et al. (2004)
z~2-3
Massive (Ha)
Blain et al. (2004)
Neri et al. (2003)
Gas rich (CO)
Progenitors of todays
massive galaxies?
All massive galaxies possibly went through
a “SCUBA phase” ~3x108 yr activity cycle
(based on gas consumption)
and then passive evolution
Faber-Jackson
Est. halo
velocity
dispersion
Swinbank et al. (2006)
Space density consistent with >3L* galaxies (duty cycle corrected)
AGN properties? Has been challenging…
An optically bright AGN (Ivison et al. 2002)
Pope et al. (2008)
(Ivison et al. 1998)
A mid-IR bright AGN
A radio-bright AGN
Alexander et al. (2005) Alexander et al. (2005)
Most moderately luminous
Most are heavily obscured
A few mid-IR/optical/radio bright AGN but most of the AGN are
X-ray faint: heavily obscured and only moderately luminous
Bolometric Luminosity typically Dominated by Star Formation
Alexander et al. (2005)
AGN contribution ~10% at
FIR and 4mJy) …indicating long (almost continuous)
black-hole growth during intense star-formation episodes
Rapid black-hole growth phase, initiated
by major mergers?
BH mass Accretion S.Form
Chapman et al. (2003)
Fuel Supply
Eddington-limited
growth
Archibald et al. (2002)
Di Matteo et al. (2005)
Eddington-limited growth during peak star formation?
Black-hole-host galaxy relationship in SMGs?
Stellar masses estimated using Spitzer IRAC (+optical+near-IR)
Borys et al. (2005)
If assumed Eddington-limited accretion then the black-hole growth
substantially lags the stellar growth (by a factor ~50!)
However:
(1) are Eddington-limited black-hole masses appropriate?
want to be able to “weigh” the black holes
(2) has the intrinsic AGN luminosity been underestimated
(extinction corrections)?
(3) are the host-galaxy masses accurate?
Weighing the Black Holes in SMGs
D.M. Alexander, AJ, submitted
Not all SMGs are heavily obscured, some have broad Ha or Hb in the
near-IR (Swinbank et al. 2004; Takata et al. 2006)
Swinbank et al. (2004)
Chapman et al. (2005)
Can “weigh” their black holes using the virial black-hole mass
estimator: MBH=G-1 RBLR V2BLR (e.g., Kaspi et al. 2000)
Eddington Ratios and Black-Hole Masses
Careful use of the virial black-hole mass estimator for the broad
Ha and Hb emission line (Greene & Ho 2005)
Spread of properties (MBH and dM/dt): For broad-line objects, median
MBH~(1-3)x108 Msolar and fEdd~0.2-0.5 (depending on BLR geometry) –
two types of broad-line SMGs: high luminosity and low luminosity
Are the Intrinsic AGN Luminosities Underestimated?
Absorption corrections consistent with other studies and AGN properties
consistent with ULIRGs (potential local analogs)
Agreement between AGN mid-IR component and intrinsic X-ray luminosity:
mid-IR appears to be isotropic indicator of AGN luminosity
Host-Galaxy Masses?
Borys et al. (2005)
Greve et al. (2005)
Stellar masses: some contaminated CO dynamical masses:
by an AGN in near-IR avg ~1011 solar masses within
(revised average ~2x1011 solar ~2 kpc radius (i.e., bulge scale)
masses with these removed)
Physical properties of SMGs used here:
• Edd rate, h>0.1 and h~0.2 (BL SMGs/obscured ULIRGs)
• This implies MBH~(0.6-1)x108 solars for typical SMGs
• M*,dyn(CO)~1011 solars for r~2kpc (within bulge; Greve et al.
2005)
• M*,stellar~2x1011 solars (Borys et al. 2005 with near-IR
excess objects removed): whole gal but ultimate system
mass?
SMGs Lie Suggestively Below Local Relationship
Consistent with Chakrabarti et al. (2007,2008)
simulations of SMGs; see talk tomorrow
And statistically below the apparent z~2 relationship
Conclusion: black-holes in typical SCUBA galaxies
appear to be comparatively small
[~(0.6-1)x108 Msolar, for ~0.1-0.2*Edd]
The black-hole growth appears to lag that of the host
galaxy in massive star-forming galaxies, in apparent
contradiction with that found for z~2 quasars/radio
galaxies
Appears to necessitate the need for an AGN
dominated phase that predominantly grows the
black hole
Major-Merger Induced Growth of Massive Galaxies?
The Dave Sanders et al. evolutionary picture
SCUBA galaxies Normal QSOs
Obscured QSOs/IR lum QSOs
Page et al. (2004)
Normal quasars: not
undergoing extreme
star formation
Page et al. (2004)
Submm quasars
Alexander et al. (2005)
Black Holes getting bigger
Testing the Evolutionary Link between Quasars
and Submm Galaxies
K. Coppin et al. in prep.
IRAM CO observations
Selected submm detected quasars in same redshift range as submm
galaxies: some are rare monsters and some are more typical systems
Comparison between Quasars and SMGs
Average gas masses and implied
CO dynamical masses similar
between SMGs and quasars (if
quasars are assumed to be more
face on: i~20 degrees)
Black-hole Host galaxy properties
Avg Quasars
Avg SMGs Quasars
Are submm Quasars at a different evolutionary stage to SMGs?
Low dyn masses consistent with other CO studies of Quasars (e.g.,
Walter+ 04) but does the CO trace the bulge in these systems?
Conclusions
• Compton-thick quasars (ID’d from optical-mid-IR spectra and
X-rays) at z~2-2.5 are as numerous as unobscured quasars
• SMGs host concordant black hole-stellar growth: all massive
galaxies were potentially SMGs at some time during the past
• The black holes of SMGs are comparatively small (typically
MBH=(0.6-1)x108 Msolar for 0.1-0.2*Edd)
• Given their host-galaxy masses (>1011 Msolar), the black hole
growth appears to lag the stellar growth, contrary to that
found in z~2 quasars
• Are submm-detected quasars more evolved than SMGs?
• the CO detected quasars have similar gas and dynamical
masses as the SMGs but have black holes ~30x larger: not
clear if the CO traces the bulge in these systems?