Towards the First Galaxies

Abstract

The formation of the first galaxies at redshifts z~10-15 signaled the transition from the simple initial state of the universe to one of ever increasing complexity. We here review recent progress in understanding their assembly process with numerical simulations, starting with cosmological initial conditions and modelling the detailed physics of star formation. In particular, we study the role of HD cooling in ionized primordial gas, the impact of UV radiation produced by the first stars, and the propagation of the supernova blast waves triggered at the end of their brief lives. We conclude by discussing promising observational diagnostics that will allow us to probe the properties of the first galaxies, such as their contribution to reionization and the chemical abundance pattern observed in extremely low-metallicity stars. Comment: 12 pages, 14 figures, appeared in "First Stars III", eds. B. O'Shea, A. Heger and T. Abel, a high resolution version (highly recommended) can be found at http://www.ita.uni-heidelberg.de/~tgreif/files/gjb07.pdf

Full text

Page 1

arXiv:0709.0102v3 [astro-ph] 20 Oct 2009
Towards the First Galaxies
Thomas H. Greif∗,†, Jarrett L. Johnson†and Volker Bromm†
∗Institut für Theoretische Astrophysik, Albert-Ueberle Strasse 2, 69120 Heidelberg, Germany
†Department of Astronomy, University of Texas, Austin, TX 78712
Abstract. The formation of the first galaxies at redshifts z ∼ 10−15 signaled the transition from the simple initial state of
the universe to one of ever increasing complexity. We here review recent progress in understanding their assembly process
with numerical simulations, starting with cosmological initial conditions and modelling the detailed physics of star formation.
In particular, we study the role of HD cooling in ionized primordial gas, the impact of UV radiation produced by the first
stars, and the propagation of the supernova blast waves triggered at the end of their brief lives. We conclude by discussing
promising observational diagnostics that will allow us to probe the properties of the first galaxies, such as their contribution
to reionization and the chemical abundance pattern observed in extremely low-metallicity stars.
Keywords: cosmology: theory — galaxies: formation — galaxies: high-redshift — H II regions — hydrodynamics — intergalactic medium
— supernovae: general
PACS: 95.30.Lz; 97.10.Bt; 97.20.Wt
INTRODUCTION
One of the key goals in modern cosmology is to study
the assembly process of the first galaxies, and under-
stand how the first stars, stellar systems, and massive
black holes formed at the end of the cosmic dark ages,
a few hundred million years after the Big Bang. With the
formation of the first stars, the so-called Population III
(Pop III), the universe was rapidly transformed into an
increasinglycomplex,hierarchicalsystem, due to the en-
ergy and heavy element input from the first stars and ac-
creting black holes [7, 17, 26, 64]. Currently, we can di-
rectly probe the state of the universe roughly a million
years after the Big Bang by detecting the anisotropies in
the cosmic microwave background (CMB), thus provid-
ing us with the initial conditions for subsequent struc-
ture formation. Complementary to the CMB observa-
tions, we can probe cosmic history all the way from the
present-day universe to roughly a billion years after the
Big Bang, using the best available ground- and space-
based telescopes. In between lies the remaining frontier,
and the first galaxies are the sign-posts of this early, for-
mative epoch.
The tight correlation between the central black hole
mass and properties of the host galactic spheroid ob-
served at low redshift indicates that galaxy formation
and massive black hole growth are fundamentally re-
lated [35, 36, 39, 96]. A popular class of models to ex-
plain this correlation invokes black hole feedback in the
form of winds or thermal energy input to regulate gas
fueling [86]. Such feedback effects might be even more
important in the shallow potential wells of high-redshift
galaxies[57]. Anopenproblemis tounderstandtheearly
growth rate of black holes, before they reach ∼ 106M⊙,
as a function of their cosmic environment, a process that
ultimately leads to the formationof quasar-sized massive
blackholesatz>6[57].Duetothe complexityinvolved,
one needs to carry out high-resolutionnumerical simula-
tions to investigate both the build-up of the first stellar
systems and the early growth of massive black holes.
The internal structure of the first galaxies involves a
wide range of physical scales, from the virial radius of
the host dark matter halo to the much smaller scales such
as the Jeans length of the cooling gas and the Bondi
radius of the first massive black hole. Furthermore, the
dynamical evolution of the first galaxies will involve
turbulence and shocks that must be resolved. Some of
the key questions are: What is the angular momentum
contentofthecoldgas,andwhat processesaffectangular
momentum transport in the first rotationally-supported
gas flow? Do global gravitational instabilities, such as
bars-within-bars, give rise to rapid transport, and what
are the consequences of rapid transport for the growth of
the central object? What is the role of local gravitational
instability and fragmentation? What is the impact of
the radiation produced by the accreting black hole on
the thermodynamic and chemical evolution of the first
galaxy? At which point will a multi-phase interstellar
medium form?
To simulate the build-up of the first stellar systems,
we have to address the feedback from the very first stars
on the surrounding intergalactic medium (IGM), and the
formation of the second generation of stars out of ma-
terial that was influenced by this feedback. There are a
numberof reasons whyaddressingthe feedbackfrom the
first stars and understanding second-generation star for-
mation is crucial:
(i) The first steps in the hierarchical build-up of struc-

Page 2

ture provide us with a simplified laboratory for studying
galaxy formation, which is one of the main outstanding
problems in cosmology, given that we have successfully
determined the parameters of the expanding background
universe. The overall strategy is to start with cosmologi-
cal initial conditions, follow the evolution up to the for-
mation of a small number (N < 10) of Pop III stars, and
trace the ensuing expansion of the supernova (SN) blast
waves after they died together with the dispersal and
mixing of the first heavy elements, towards the forma-
tion of second-generation stars out of enriched material
[42, 101].
(ii) The initial burst of Pop III star formation may have
been rather brief due to the strong negative feedback ef-
fects that likely acted to self-limit this formation mode
[41, 104]. Second-generation star formation, therefore,
might well have been cosmologically dominant com-
pared to Pop III stars. Despite their importance for cos-
mic evolution, e.g., by possibly constituting the majority
of sources for the initial stages of reionization at z > 10,
we currently do not know the properties, and most im-
portantlythetypicalmassscale, ofthesecond-generation
stars that formed in the wake of the very first stars.
(iii) A subset of second-generation stars, those with
masses below ≃ 1 M⊙, would have survived to the
present day. Surveys of extremely metal-poor Galactic
halo stars therefore provide an indirect window into the
Pop III era by scrutinizing their chemical abundancepat-
terns, which reflect the enrichment from a single, or at
most a small multiple of, Pop III SNe [9, 37]. Stellar ar-
chaeologythus provides unique empirical constraints for
numerical simulations, from which one can derive theo-
retical abundance patterns to be compared with the data.
Recently, observations of the highest-redshift quasars
andgalaxiesto datehaveopenedupatantalizingwindow
into the state of the universe at z > 7 [8, 13, 32, 48, 55,
76, 80, 92, 103]. Studying the nature of star formation
at even higher redshifts is of great interest, because the
systems that we can directly observeat z>7 will already
showthesignatureofpreviousstars. Anintriguingexam-
ple is the recently discovered J−band dropout HUDF-
JD2, found through deep HST/VLT/Spitzer imaging,
and interpreted to be a massive (∼ 6×1011M⊙) post-
starburst galaxy at z > 6.5 [65]. Since the observed stel-
lar population in HUDF-JD2 is old, star formation could
have begun already at z ∼ 15, and this galaxy could
thus contain the signature of first- and second-generation
stars.
Existing and planned observatories, such as HST,
Keck, VLT, and the James Webb Space Telescope
(JWST), planned for launch around 2013, yield data
on stars and quasars less than a billion years after the
Big Bang. The ongoing Swift gamma-ray burst (GRB)
mission providesus with a possible windowinto massive
star formation at the highest redshifts [18, 22, 56]. Mea-
surements of the near-IR cosmic background radiation,
both in terms of the spectral energy distribution and
the angular fluctuations provide additional constraints
on the overall energy production due to the first stars
[29, 33, 52, 62, 79]. Understanding the formation of the
first galaxies is thus of great interest to observational
studies conducted both at high redshifts and in our local
Galactic neighborhood.
ASSEMBLY OF THE FIRST GALAXIES
How massive were the first galaxies, and when did they
emerge? Theory predicts that dark matter (DM) halos
containing a mass of ∼ 108M⊙and collapsing at z ∼
10−15 were the hosts for the first bona fide galaxies.
These dwarf systems are special in that their associated
virial temperature exceeds the threshold, ∼ 104K, for
cooling due to atomic hydrogen [70]. These so-called
‘atomic-cooling halos’ did not rely on the presence of
molecular hydrogen to enable cooling of the primordial
gas. In addition, their potential wells were sufficiently
deep to retain photoionization heated gas, as well as SN
shocked gas, in contrast to the shallow potential wells of
minihalos [28, 42, 61, 66]. These are arguably minimum
requirementstoset upa self-regulatedprocessofstarfor-
mation that comprises more than one generation of stars,
and is embeddedin a multi-phase interstellar medium. In
the following, we discuss some of the key processes that
govern the assembly of these first dwarf galaxies.
To avoid confusion, we would like to comment on
a change in terminology. It has become evident that
Pop III star formation might actually consist of two dis-
tinct modes: one where the primordial gas collapses into
a DM minihalo (see below), and one where the metal-
free gas becomes significantly ionized prior to the onset
of gravitational runaway collapse [49]. We had termed
this latter mode of primordial star formation ‘Pop II.5’
[41, 49, 60]. To more clearly indicate that both modes
pertain to metal-free star formation, we here follow the
new classification scheme suggested by Chris McKee
(see McKee in these proceedings). Within this scheme,
the minihalo Pop III mode is now termed Pop III.1,
whereas the second mode (formerly ‘Pop II.5’) is now
called Pop III.2. The hope is that McKee’s terminology
will gain wide acceptance.
Role of HD Cooling
While the very first Pop III stars (so-called Pop III.1),
with masses of the order of 100 M⊙, formed within DM
minihalosinwhichprimordialgascoolsbyH2molecules
alone, the HD molecule can play an important role in

Page 3

the cooling of primordialgas in several situations, allow-
ing the temperature to drop well below 200 K [1, 14]. In
turn, this efficient cooling may lead to the formation of
primordial stars with masses of the order of 10 M⊙(so-
called Pop III.2) [49]. In general, the formation of HD,
and the concomitant cooling that it provides, is found
to occur efficiently in primordial gas which is strongly
ionized, owing ultimately to the high abundance of elec-
trons which serve as catalyst for molecule formation in
the early universe [84].
Efficient cooling by HD can be triggered within the
relic H II regions that surround Pop III.1 stars at the
end of their brief lifetimes, owing to the high electron
fractionthatpersists in the gasas it cools andrecombines
[51,67,105].TheefficientformationofHD canalso take
place when the primordial gas is collisionally ionized,
such as behind the shocks driven by the first SNe or in
the virialization of massive DM halos [41, 49, 59, 85].
In Figure 1, we show the HD fraction in primordial gas
in four distinct situations: within a minihalo in which
the gas is never strongly ionized, behind a 100 km s−1
shock wave driven by a SN, in the virialization of a
3σ DM halo at redshift z = 15, and in the relic H II
region generated by a Pop III.1 star at z ∼ 20 [49]. Also
shown is the critical HD fraction necessary to allow the
primordial gas to cool to the temperature floor set by the
CMB at these redshifts. Except for the situation of the
gas in the virtually un-ionized minihalo, the fraction of
HD becomes large quickly enough to play an important
role in the cooling of the gas, allowing the formation of
Pop III.2 stars.
Figure 2 schematically shows the characteristic
masses of the various stellar populations that form in
the early universe. In the wake of Pop III.1 stars formed
in DM minihalos, Pop III.2 star formation ensues in
regions which have been previously ionized, typically
associated with relic H II regions left over from massive
Pop III.1 stars collapsing to black holes, while even later,
when the primordial gas is locally enriched with metals,
Pop II stars begin to form [20, 41]. Recent simulations
confirm this picture, as Pop III.2 star formation ensues
in relic H II regions in well under a Hubble time, while
the formation of Pop II stars after the first SN explosions
is delayed by more than a Hubble time [42, 105, 106].
Radiative Feedback
Due to their extreme mass scale, Pop III.1 stars emit
copiousamountsofionizingradiation,as well as astrong
flux of H2-dissociating Lyman-Werner (LW) radiation
[16, 81]. Thus, the radiation from the first stars dramati-
cally influences their surroundings, heating and ionizing
thegaswithina fewkiloparsec(physical)aroundthepro-
FIGURE 1.
mordial gas which cools in four distinct situations. The solid
line corresponds to gas with an initial density of 100 cm−3,
which is compressed and heated by a SN shock with velocity
vsh= 100 km s−1at z = 20. The dotted line corresponds to
gas at an initial density of 0.1 cm−3shocked during the for-
mation of a 3σ halo at z = 15. The dashed line corresponds
to unshocked, un-ionized primordial gas with an initial density
of 0.3 cm−3collapsing inside a minihalo at z = 20. Finally,
the dash-dotted line shows the HD fraction in primordial gas
collapsing from an initial density of 0.3 cm−3inside a relic
H II region at z = 20. The horizontal line at the top denotes
the cosmic abundance of deuterium. Primordial gas with an
HD abundance above the critical value, XHD,crit, denoted by the
bold dashed line, can cool to the CMB temperature within a
Hubble time.
Evolution of the HD abundance, XHD, in pri-
FIGURE 2.
redshift. Pop III.1 stars, formed from unshocked, un-ionized
primordial gas are characterized by masses of the order of
100 M⊙.Pop II stars,formed ingaswhich isenriched withmet-
als, emerged at lower redshifts and have characteristic masses
of the order of 1 M⊙. Pop III.2 stars, formed from ionized
primordial gas, have characteristic masses reflecting the fact
that they form from gas that has cooled to the temperature of
the CMB. Thus, the characteristic mass of Pop III.2 stars is a
function of redshift, but is typically of the order of 10 M⊙.
Characteristic stellar mass as a function of

Page 4

genitor, and destroying the H2and HD molecules locally
within somewhat larger regions [2, 5, 34, 51, 54, 98].
Additionally, the LW radiation emitted by the first stars
could propagate across cosmological distances, allow-
ing the build-up of a pervasive LW backgroundradiation
field [43].
The impact of radiation from the first stars on their lo-
cal surroundingshas importantimplications for the num-
bers and types of Pop III stars that form. The photo-
heating of gas in the minihalos hosting Pop III.1 stars
drives strong outflows, lowering the density of the pri-
mordial gas and delaying subsequent star formation by
up to 100 Myr [51, 98, 105]. Furthermore, neighboring
minihalos may be photoevaporated,delaying star forma-
tion in such systems as well [4, 42, 83, 93, 99]. The pho-
todissociation of molecules by LW photonsemitted from
local star-forming regions will, in general, act to delay
star formation by destroying the main coolants that al-
low the gas to collapse and form stars.
The photoionization of primordial gas, however, can
ultimately lead to the production of copious amounts of
molecules within the relic H II regions surrounding the
remnants of Pop III.1 stars [50, 67, 70, 75]. Recent simu-
lations tracking the formation of, and radiative feedback
from, individual Pop III.1 stars in the early stages of the
assembly of the first galaxies have demonstrated that the
accumulation of relic H II regions has two important ef-
fects. First, the HD abundance that develops in relic H II
regions allows the primordial gas to re-collapse and cool
to the temperature of the CMB, possibly leading to the
formation of Pop III.2 stars in these regions [51, 106].
Second, the molecule abundance in relic H II regions,
along with their increasing volume-filling fraction, leads
to a large optical depth to LW photons over physical dis-
tances of the order of several kiloparsecs. The develop-
ment of a high optical depth to LW photons over such
short length-scales suggests that the optical depth to LW
photons over cosmological scales may be very high, act-
ing to suppress the build-up of a background LW radia-
tion field, and mitigating negative feedback on star for-
mation.
Figure 3 shows the chemical composition of primor-
dial gas in relic H II regions, in which the formation of
H2molecules is catalyzed by the high residual electron
fraction. Figure 4 shows the average optical depth to LW
photons across the simulation box, which rises with time
owing to the increasing number of relic H II regions.
Massive Black Hole Growth
The observations of quasars at redshift z > 6 in the
SloanDigital Sky Survey(SDSS)suggest that some black
holes in the early universe grew to have masses in excess
FIGURE 3.
While all molecules are destroyed in and around active H II
regions, the high residual electron fraction in relic H II re-
gions catalyzes the formation of an abundance of H2and HD
molecules. The light and dark shades of blue denote regions
with a free electron fraction of 5×10−3and 5×10−4, respec-
tively, while the shades of green denote regions with an H2
fraction of 10−4, 10−5, and 3×10−6, in order of decreasing
brightness. The regions with the highest molecule abundances
lie within relic H II regions, which thus play an important role
in subsequent star formation, allowing molecules to become
shielded from photodissociating radiation and altering thecool-
ing properties of the primordial gas.
The chemical interplay in relic H II regions.
of 109M⊙ within the first billion years after the Big
Bang [30, 31]. Many massive Pop III stars may have
collapsed directly to form black holes at the end of
their brief lifetimes, thus providing the seed black holes
which then accreted gas and grew to be the supermassive
black holes powering the high-redshift quasars that are
observed today [45].
In order for Pop III remnant black holes to accrete
∼ 109M⊙of mass by z ∼ 6, these objects would have
to accrete gas at or near the Eddington rate for hundreds
of millions of years. In turn, these black holes can only
accrete at such high rates if the surrounding density ex-
ceeds102cm−3(see Figure5)[50].However,as Figure6
shows, the photoheating of the gas surrounding the first
Pop III stars in minihalos drives strong outflows from the
central star, leaving a remnant black hole to reside in the
middleof the evacuatedminihalowith densities typically
well below 1 cm−3. Thus, the high accretion rates neces-
sary to explain the presence of 109M⊙black holes at
z ∼6 are not possible for at least 50 Myr after the forma-
tion of the black hole, at which time DM halo mergers
and the re-collapse of the photoheated gas may raise the
central density of the gas to above 102cm−3[51]. Sim-
ulations tracking the later accretion history of such seed
black holes have also found that it is difficult for these
objects to grow fast enough to explain the SDSS quasar
observations (see also Alvarez et al. in these proceed-
ings) [57, 74]. Therefore, more exotic origins for these

Page 5

FIGURE 4.
different scales, as a function of redshift. The diamonds denote
the optical depth averaged over the entire cosmological box
of comoving length 660 kpc, while the plus signs denote the
optical depth averaged over a cube of 220 kpc (comoving) per
side, centered on the middle of the box. The solid line denotes
the average optical depth that would be expected for a constant
H2fraction of 2×10−6(primordial gas), which changes only
due to cosmic expansion.
Optical depth to LW photons averaged over two
FIGURE 5.
a function of time, assuming the black hole accretes gas at
a temperature of 200 K and at constant density. The solid,
dotted, dashed, and dot-dashed lines show cases with densities
of 0.1 cm−3, 1 cm−3, 10 cm−3, and 100 cm−3, respectively.
The triple dot-dashed line shows the mass of the black hole as
a function of time, assuming that it accretes at the Eddington
limit. Clearly, for the black hole to begin accreting at such a
high rate while its mass is still of the order of 100 M⊙, the
surrounding density must exceed 102cm−3.
The mass of an initially 100 M⊙black hole as
supermassive black holes may be required [10, 19, 58].
FIGURE 6.
primordial gas heated and ionized by radiation from a 100 M⊙
and 200 M⊙primordial star after their main-sequence lifetimes
of 3 Myr and 2 Myr, respectively. The central density drops to
well below 1 cm−3, which is insuffiecient to enable subsequent
accretion at the Eddington rate (see Figure 5).
The density, temperature and radial velocity of
The First Supernova Explosions
Recent numerical simulations have indicated that pri-
mordial stars forming in DM minihalos typically attain
100M⊙by efficientaccretion,andmight evenbecomeas
massive as 500 M⊙[21, 71, 72, 107]. After their main-
sequence lifetimes of typically 2 − 3 Myr, stars with
masses below ≃ 100 M⊙are thought to collapse directly
to black holes without significant metal ejection, while
in the range 140−260 M⊙a pair-instability supernova
(PISN) disrupts the entire progenitor, with explosion en-
ergies ranging from 1051−1053ergs, and yields up to
0.5 [45, 46]. Less massive primordial stars with a high
degree of angular momentum might explode with simi-
lar energies as hypernovae[94, 97].
The significant mechanical and chemical feedback ef-
fects exerted by such explosions have been investigated
with a numberof detailed calculations, but these were ei-
ther performed in one dimension [53, 59, 77], or did not
start from realistic initial conditions [23, 69]. The most
realistic simulation to date took cosmologicalinitial con-
ditions into account, and followed the evolution of the
gas until the formation of the first minihalo at z ≃ 20
(see Figure 7) [42]. After the gas approached the ‘loiter-
ing regime’ at nH≃ 104cm−3, the formation of a pri-
mordial star was assumed, and a photoheating and ray-
tracing routine determined the size and structure of the
resulting H II region (see Figure 8) [51]. An explosion
energy of 1052ergs was then injected as thermal energy

Page 6

FIGURE 7.
lineof sight inasliceof 14 kpc (comoving) around thefirststar,
forming in a halo of total mass Mvir≃ 5×105M⊙at z ≃ 20.
Evidently, the host halo is part of a larger group of less massive
minihalos, and subject to the typical bottom-up evolution of
structure formation.
Hydrogen number density averaged along the
into a small region around the progenitor, and the sub-
sequent expansion of the SN remnant was followed until
the blast wave effectively dissolved into the IGM. The
cooling mechanisms responsible for radiating away the
energy of the SN remnant, the temporal behavior of the
shock, and its morphology could thus be investigated in
great detail (see Figure 9).
Dynamical Evolution
In analogy to present-day SNe, the evolution of
SN remnants in the early universe can be decom-
posed into four physically distinct stages: free expan-
sion(FE),Sedov-Taylorblast wave(ST),pressure-driven
snowplow (PDS), and momentum-conservingsnowplow
(MCS) [73]. This allows the introduction of a simple an-
alytic model, which reproduces the simulation results of
[42] relatively well (see Figure 9). It was found that the
SN remnant propagates for a Hubble time at z ≃ 20 to a
final mass-weightedmean shock radius of 2.5 kpc (phys-
ical), roughly half the size of the H II region, and sweeps
up a total (gas) mass of 2.5×105M⊙. The radial disper-
sion of the SN remnant increased dramatically once the
shock left the host halo, although the bulk of the swept-
up gas was expelled into the voids of the general IGM.
FIGURE 8.
a slice of 14 kpc (comoving) around the star after its main
sequence lifetime of 2 Myr. Ionizing radiation has penetrated
nearby minihalos and extends up to 5 kpc (physical) around
the source, heating the IGM to roughly 2×104K, while some
high-density regions have effectively shielded themselves.
Temperature averaged along the line of sight in
Mechanical Feedback
As confirmed in [42], highly energetic SN explosions
are sufficient to entirely disrupt the host halo and evac-
uate their gaseous content [60]. This is directly related
to their shallow potential wells, and was also found in
previouscalculations[23, 53].Subsequentstar formation
does not ensue for at least a Hubble time at z ≃ 20, and
in the specific scenario considered, the first Pop II stars
are expected to form out of enriched material at z ≃ 10
[42].
Additional simulations in the absence of a SN explo-
sion were performed to investigate the effect of photo-
heating and the impact of the SN shock on neighbour-
ing minihalos (see Figure 10) [42]. For the case dis-
cussed in this work, the SN remnant exerted positive me-
chanical feedback on neighbouring minihalos by shock-
compressing their cores, while photoheating marginally
delayed star formation. Although a viable theoretical
possibility, secondary star formation in the dense shell
via gravitational fragmentation was not observed, pri-
marily due to the previous photoheating by the progen-
itor and the rapid adiabatic expansion of the post-shock
gas [59, 60, 77].

Page 7

FIGURE 9.
(a)), and swept-up (gas) mass (Panel (b)) of a 1052ergs SN
explosion in the high-redshift universe as a function of time
(black dots), compared to the analytic model (dashed line). In
the specific scenario considered, a final mass-weighted mean
shock radius of 2.5 kpc (physical) and a final swept-up mass of
2.5×105M⊙was found. The shaded region shows the radial
dispersion of theshock, indicating that itincreases dramatically
once the SN remnant leaves the host halo and encounters the
first neighbouring minihalos.
The mass-weighted mean shock radius (Panel
FIGURE 10.
progenitor for all star-forming minihalos affected by the SN
shock. The shades of the symbols indicate their affiliation,
i.e. black, dark grey and light grey symbols represent the no-
feedback, photoheating-only and main simulation runs, respec-
tively, while the shapes of the symbols denote the individual
halos. For orientation, thedashed lineshows themass-weighted
mean shock radius at late times according to Figure 9. In the
cosmological realizationconsidered, photoheating significantly
delays star formation, while the SN shock compresses gas in
neighbouring minihalos and slightly accelerates their collapse.
The collapse times and distances from the SN
FIGURE 11.
line of sight in a slice of 14 kpc (comoving), overlayed with the
distribution of all metal particles (bright orange) after 200 Myr,
when the shock stalls. Most metals are dispersed into the voids
of the general IGM, since the metal-rich interior expands adia-
batically into the cavities created by the shock.
Hydrogen number density averaged along the
Distribution of Metals
The dispersal of metals by the first SN explosions
transformed the IGM from a simple primordial gas to a
highly complex medium in terms of chemistry and cool-
ing, which ultimately enabled the formation of the first
low-mass stars. However, this transition required at least
a Hubble time, since the presence of metals became im-
portant only after the SN remnant had stalled and the en-
riched gas re-collapsed to high densities [42]. Further-
more, the metal distribution was highly anisotropic, as
the post-shock gas expanded into the voids in the shape
of an ‘hour-glass’, with a maximum extent similar to the
final mass-weighted mean shock radius (see Figure 11)
[42].
To efficiently mix the metals with all components of
the swept-up gas, a DM halo of at least Mvir≃ 108M⊙
had to be assembled [42], and with an initial yield of 0.1,
the average metallicity of such a system would accumu-
late to Z ≃ 10−2.5Z⊙, well above any critical metallic-
ity [15, 20, 82]. Thus, if energetic SNe were a common
fate for the first stars, they would have deposited metals
on large scales before massive galaxies formed and out-
flows were suppressed. Hints to such ubiquitous metal
enrichment have been found in the low column density
Lyα forest [3, 87, 88], and in dwarf spheroidal satellites
of the Milky Way [47].

Page 8

OBSERVATIONAL CONSTRAINTS
An increasing number of direct and indirect observa-
tions have become feasible during the last few years,
due to rapid progress in observational methods and tech-
niques. The most prominent among these concern the
CMB optical depth to Thomson scattering, the near-IR
background, high-redshift GRBs, and the possibility of
scrutinizing the nature of the first stars by metals found
in the oldest Galactic halo stars, dubbed ‘stellar arche-
ology’. We here briefly discuss these promising obser-
vational avenues, and elaborate on their implications for
the first stars and galaxies.
Optical Depth to Thomson Scattering
Oneofthemostimportantconstraintsonearlystarfor-
mation is the CMB optical depth to Thomson scattering,
recently revised to τ ≃ 0.09±0.03 after the Wilkinson
Microwave Anisotropy Probe (WMAP) three-year data
release [89]. Combined with the absence of the Gunn-
Peterson trough in the spectra of high-redshift quasars,
this measurement provides an integral constraint on the
total ionizing photon production at z > 6 [6, 102].
In recent work, the contribution of the first stars to the
ionizing photon budget was determined by semianalytic
star formation rates based on the Sheth-Tormen formal-
ism [41]. It was found that a top-heavy primordial popu-
lation (consisting of Pop III.1 stars) must be terminated
fairly rapidly in order to not overproduce the observed
optical depth (see Figure 12). Such efficient feedback is
possible via photoionization heating, which delays and
possibly changes the mode of star formation in neigh-
bouring minihalos, depending on the state of the mini-
halo collapse [4, 63, 93, 99, 105].
Near-Infrared Background
Another observation with the potential to yield valu-
able information on the nature of the first ionizing
sources is the detection of an excess in the near-IR back-
ground of the order of 1 nW m−2sr−1in the wave-
length regime 1−2 µm [24]. If a fraction of this light
is contributed by early stars, it corresponds to redshifted
Lyα photons emitted between z ≃ 7 and z ≃ 15. Ap-
plying a semianalytic model of primordial star forma-
tion and deriving the total luminosity, it has been found
that Pop III.1 and Pop III.2 combined must contribute at
a level less than 10−3nW m−2sr−1in order to avoid
premature reionization [41]. To determine what fraction
is contributed by metal-free stars, one must analyze the
fluctuation power of the IR background at scales small
enoughto capture their strong clustering [29, 52, 62, 79].
FIGURE 12.
optical depths (Panel (b)) for a weak (dotted line), intermediate
(dashed line), and strong (dot-dashed line) Pop III.1 mode. For
comparison, Panel (b) includes the range of allowed values for
the optical depth (long-dashed lines).
Ionization histories (Panel (a)), and implied
High-redshift Gamma-ray Bursts
GRBs are believed to originate in compact remnants
(neutron stars or black holes) of massive stars, and their
high luminosities make them detectable out to the edge
of the visible universe [27, 56]. GRBs offer the opportu-
nity to detect the most distant (and hence earliest) pop-
ulation of massive stars. In the hierarchical assembly
process of DM halos, the first galaxies should have had
lower masses (and lower stellar luminosities) than their
low-redshift counterparts. Consequently,the characteris-
tic luminosity of galaxies or quasars is expected to de-
cline with increasingredshift.GRB afterglows,whichal-
ready produce a peak flux comparable to that of quasars
or starburst galaxies at z ∼ 1−2, are therefore expected
to outshine any competing source at the highest red-
shifts. As the electromagnetically-brightest explosions
in the universe, GRBs should be detectable out to red-
shifts z > 10 [27, 56]. High-redshift GRBs can be iden-
tified through IR photometry, based on the Lyα break
induced by absorption of their spectrum at wavelengths
below 1.216 µm [(1+z)/10].Follow-upspectroscopyof
high-redshift candidates can then be performed on a 10-
meter-class telescope. Recently, the ongoing Swift mis-
sion has detected a GRB originating at z ≃ 6.3 [40, 44],
thusdemonstratingtheviabilityofGRBs as probesofthe
early universe.

Page 9

Although the nature of the central engine that pow-
ers the relativistic jets of GRBs is still unknown, recent
evidenceindicates that long-durationGRBs tracethe for-
mation of massive stars [11, 12, 68, 95, 100], and in par-
ticular that long-durationGRBs are associated with Type
Ib/c SNe [91]. Since the first stars in the universe are
predicted to be predominantly massive [1, 14, 17], their
death might give rise to large numbers of GRBs at high
redshifts. In contrast to quasars of comparable bright-
ness,GRB afterglowsareshort-livedandrelease∼10or-
ders of magnitude less energyinto the surroundingIGM.
Beyondthescaleoftheirhostgalaxy,theyhaveanegligi-
ble effect on their cosmological environment. However,
the feedback from a single GRB or SN on the gas con-
fined within the first galaxies could be dramatic, since
the binding energy of most galaxies at z > 10 is lower
than 1051ergs[7]. Consequently,they are ideal probes of
the IGM during the reionization epoch. Their rest-frame
UV spectra can be used to probe the ionization state of
the IGMthroughthe spectralshapeofthe Gunn-Peterson
(Lyα) absorption trough, or its metal enrichment history
throughthe intersectionof enrichedbubbles of SN ejecta
from early galaxies [38].
What is thesignatureofGRBs thatoriginatesinmetal-
free, Pop III progenitors? Simply knowing that a given
GRB camefroma highredshiftisnotsufficienttoreacha
definite conclusion as to the nature of the progenitor. For
example, the currently highest-redshift GRB at z ≃ 6.3
clearly did not originate from a Pop III progenitor, given
the inferred level of metal enrichment in the host sys-
tem of a few percent solar [25]. Pregalactic metal en-
richment was likely quite inhomogeneous, and we ex-
pect normal Pop I and II stars to exist in galaxies that
were already metal-enriched at these high redshifts [22].
Pop III and Pop I/II star formation is thus predicted to
haveoccurredconcurrentlyat z>5. Howis thepredicted
high mass-scale for Pop III stars reflected in the obser-
vational signature of the resulting GRBs? Preliminary
results indicate that circumburst densities are systemat-
ically higher in Pop III environments. GRB afterglows
will then be much brighter than for conventional GRBs.
In addition, due to the systematically increased progen-
itor masses, the Pop III distribution may be biased to-
wards long-durationevents. Figure 13 leads to the robust
expectation that ∼ 10% of all Swift bursts should origi-
nate at z > 5. This prediction is based on the contribu-
tion from Pop I/II stars which are known to exist even at
these high redshifts. Additional GRBs could be triggered
by Pop III stars, with a highly uncertain efficiency.
Stellar Archeology
The discovery of extremely metal-poor stars in the
Galactic halo has made studies of the chemical com-
FIGURE 13.
[22]. Shown is the observed number of bursts per year,
dNobs
GRB/dln(1 + z), as a function of redshift. All rates are
calculated with a constant GRB efficiency, ηGRB≃ 2 ×
10−9bursts M−1
⊙. Dotted lines: Contribution to the observed
GRB rate from Pop I/II and Pop III for the case of weak chemi-
cal feedback. Dashed lines: Contribution to the GRB rate from
Pop I/II and Pop III for the case of strong chemical feedback.
Filled circle: GRB rate from Pop III stars if these were respon-
sible for reionizing the universe at z ∼ 17.
Predicted GRB rate to be observed by Swift
position of low-mass Pop II stars powerful probes of
the conditions in which the first low-mass stars formed.
While it is widely accepted that metals are required for
the formation of low-mass stars, two general classes of
competing models for the Pop III – Pop II transition
are discussed in the literature: (i) atomic fine-structure
line cooling [20, 78]; and (ii) dust-induced fragmenta-
tion [82]. Within the fine-structure model, C II and O I
have been suggested as main coolants [20], such that
low-mass star formation can occur in gas that is enriched
beyondcritical abundancesof [C/H]crit≃−3.5±0.1and
[O/H]crit≃ −3 ± 0.2. The dust-cooling model, on the
other hand,predicts critical abundancesthat are typically
smaller by a factor of 10−100.
Based on the theory of atomic line cooling [20], a
new function, the ‘transition discriminant’ has been in-
troduced:
10[C/H]+0.3×10[O/H]?
Dtrans≡ log10
?
,(1)
such that low-mass star formation requires Dtrans >
Dtrans,crit≃ −3.5±0.2 [37]. Figure 14 shows values of
Dtrans for a large number of the most metal-poor stars
available in the literature. While theories based on dust
cooling can be pushed to accommodate the lack of stars
with Dtrans<Dtrans,crit, it appears that the atomic-cooling

Page 10

FIGURE 14.
stars collected from the literature as a function of [Fe/H]. Top
panel: Galactic halo stars. Bottom panel: Stars in dSph galaxies
and globular clusters. G indicates giants, SG subgiants. The
criticallimitismarked withadashed line.Thedotted linesrefer
to the uncertainty. The detailed references for the various data
sets can be found in [37].
Transition discriminant, Dtrans, for metal-poor
theory for the Pop III – Pop II transition naturally ex-
plains the existing data on metal-poor stars. Future sur-
veys of Galactic halo stars will allow to further populate
plots such as Figure 14, and will providevaluable insight
into the conditionsof the early universein which the first
low-mass stars formed.
OUTLOOK
Understanding the formation of the first galaxies marks
the frontierof high-redshiftstructure formation.It is cru-
cial to predict their properties in order to develop the op-
timal searchandsurveystrategiesforthe JWST. Whereas
ab-initio simulations of the very first stars can be carried
out from first principles, and with virtually no free pa-
rameters,onefacesamuchmoredauntingchallengewith
thefirstgalaxies.Now,theprevioushistoryofstarforma-
tion has to be considered, leading to enhanced complex-
ity in the assembly of the first galaxies. One by one, all
the complex astrophysical processes that play a role in
more recent galaxy formation appear back on the scene.
Among them are external radiation fields, comprising
UV and X-ray photons, and possibly cosmic rays pro-
duced in the wake of the first SNe [90]. There will be
metal-enriched pockets of gas which could be pervaded
by dynamically non-negligible magnetic fields, together
withturbulentvelocityfields builtupduringthevirializa-
tion process. However, the goal of making useful predic-
tions for the first galaxies is now clearly drawing within
reach, and the pace of progress is likely to be rapid.
ACKNOWLEDGMENTS
V. B. acknowledges support from NSF grant AST-
0708795and NASA Swift grantNNX07AJ636.The sim-
ulations presented here were carried out at the Texas Ad-
vanced Computing Center (TACC). We are grateful to
Paul Navrátil and Karla Vega at TACC for help with vi-
sualizations. T. H. G. would like to thank Paul Clark for
suggestions which have improved the layout and content
of this work.
REFERENCES
1. Abel, T., Bryan, G. L., & Norman, M. L. 2002, Science,
295, 93
Abel, T., Wise, J. H., & Bryan, G. L. 2007, ApJ, 659,
L87
Aguirre, A., Schaye, J., Hernquist, L., Kay, S., Springel,
V., & Theuns, T. 2005, ApJ, 620, L13
Ahn, K., & Shapiro, P. R. 2007, MNRAS, 375, 881
Alvarez, M. A., Bromm, V., & Shapiro, P. R. 2006, ApJ,
639, 621
Alvarez, M. A., Shapiro, P. R., Ahn, K., & Iliev, I. T.
2006, ApJ, 644, L101
Barkana, R., & Loeb, A. 2001, Phys. Rep., 349, 125
Barton, E. J., Davé, R., Smith, J.-D. T., Papovich, C.,
Hernquist, L., & Springel, V. 2004, ApJ, 604, L1
Beers, T. C., & Christlieb, N. 2005, ARA&A, 43, 531
Begelman, M. C., Volonteri, M., & Rees, M. J. 2006,
MNRAS, 370, 289
Blain, A. W., & Natarajan, P. 2000, MNRAS, 312, L35
Bloom, J. S., Kulkarni, S. R., & Djorgovski, S. G. 2002,
AJ, 123, 1111
Bouwens, R. J., & Illingworth, G. D. 2006, Nature, 443,
189
Bromm, V., Coppi, P. S., & Larson, R. B. 2002, ApJ,
564, 23
Bromm, V., Ferrara, A., Coppi, P. S., & Larson, R. B.
2001, MNRAS, 328, 969
Bromm, V., Kudritzki, R. P., & Loeb, A. 2001, ApJ, 552,
464
Bromm, V., & Larson, R. B. 2004, ARA&A, 42, 79
Bromm, V., & Loeb, A. 2002, ApJ, 575, 111
—. 2003, ApJ, 596, 34
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.

Page 11

20.
21.
22.
23.
—. 2003, Nature, 425, 812
—. 2004, New Astronomy, 9, 353
—. 2006, ApJ, 642, 382
Bromm, V., Yoshida, N., & Hernquist, L. 2003, ApJ,
596, L135
Cambrésy, L., Reach, W. T., Beichman, C. A., & Jarrett,
T. H. 2001, ApJ, 555, 563
Campana, S., Lazzati, D., Ripamonti, E., Perna, R.,
Covino, S., Tagliaferri, G., Moretti, A., Romano, P.,
Cusumano, G., & Chincarini, G. 2007, ApJ, 654, L17
Ciardi, B., & Ferrara, A. 2005, Space Science Reviews,
116, 625
Ciardi, B., & Loeb, A. 2000, ApJ, 540, 687
Dijkstra, M., Haiman, Z., Rees, M. J., & Weinberg,
D. H. 2004, ApJ, 601, 666
Dwek, E., Arendt, R. G., & Krennrich, F. 2005, ApJ,
635, 784
Fan, X., Hennawi, J. F., Richards, G. T., Strauss, M. A.,
Schneider, D. P., Donley, J. L., Young, J. E., Annis, J.,
Lin, H., Lampeitl, H., Lupton, R. H., Gunn, J. E., Knapp,
G. R., Brandt, W. N., Anderson, S., Bahcall, N. A.,
Brinkmann, J., Brunner, R. J., Fukugita, M., Szalay,
A. S., Szokoly, G. P., & York, D. G. 2004, AJ, 128, 515
Fan, X., Strauss, M. A., Richards, G. T., Hennawi,
J. F., Becker, R. H., White, R. L., Diamond-Stanic,
A. M., Donley, J. L., Jiang, L., Kim, J. S., Vestergaard,
M., Young, J. E., Gunn, J. E., Lupton, R. H., Knapp,
G. R., Schneider, D. P., Brandt, W. N., Bahcall,
N. A., Barentine, J. C., Brinkmann, J., Brewington,
H. J., Fukugita, M., Harvanek, M., Kleinman, S. J.,
Krzesinski, J., Long, D., Neilsen, Jr., E. H., Nitta, A.,
Snedden, S. A., & Voges, W. 2006, AJ, 131, 1203
Fan, X., Strauss, M. A., Schneider, D. P., Becker, R. H.,
White, R. L., Haiman, Z., Gregg, M., Pentericci, L.,
Grebel, E. K., Narayanan, V. K., Loh, Y.-S., Richards,
G. T., Gunn, J. E., Lupton, R. H., Knapp, G. R., Ivezi´ c,
Ž., Brandt, W. N., Collinge, M., Hao, L., Harbeck,
D., Prada, F., Schaye, J., Strateva, I., Zakamska, N.,
Anderson, S., Brinkmann, J., Bahcall, N. A., Lamb,
D. Q., Okamura, S., Szalay, A., & York, D. G. 2003, AJ,
125, 1649
Fernandez, E. R., & Komatsu, E. 2006, ApJ, 646, 703
Ferrara, A. 1998, ApJ, 499, L17+
Ferrarese, L., & Ford, H. 2005, Space Science Reviews,
116, 523
Ferrarese, L., & Merritt, D. 2000, ApJ, 539, L9
Frebel, A., Johnson, J. L., & Bromm, V. 2007, MNRAS,
380, L40
Furlanetto, S. R., & Loeb, A. 2003, ApJ, 588, 18
Gebhardt, K., Bender, R., Bower, G., Dressler, A.,
Faber, S. M., Filippenko, A. V., Green, R., Grillmair,
C., Ho, L. C., Kormendy, J., Lauer, T. R., Magorrian, J.,
Pinkney, J., Richstone, D., & Tremaine, S. 2000, ApJ,
539, L13
Gehrels, N., Chincarini, G., Giommi, P., Mason, K. O.,
Nousek, J. A., Wells, A. A., White, N. E., Barthelmy,
S. D., Burrows, D. N., Cominsky, L. R., Hurley,
K. C., Marshall, F. E., Mészáros, P., Roming, P. W. A.,
Angelini, L., Barbier, L. M., Belloni, T., Campana,
S., Caraveo, P. A., Chester, M. M., Citterio, O., Cline,
T. L., Cropper, M. S., Cummings, J. R., Dean, A. J.,
Feigelson, E. D., Fenimore, E. E., Frail, D. A., Fruchter,
A. S., Garmire, G. P., Gendreau, K., Ghisellini, G.,
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
Greiner, J., Hill, J. E., Hunsberger, S. D., Krimm, H. A.,
Kulkarni, S. R., Kumar, P., Lebrun, F., Lloyd-Ronning,
N. M., Markwardt, C. B., Mattson, B. J., Mushotzky,
R. F., Norris, J. P., Osborne, J., Paczynski, B., Palmer,
D. M., Park, H.-S., Parsons, A. M., Paul, J., Rees, M. J.,
Reynolds, C. S., Rhoads, J. E., Sasseen, T. P., Schaefer,
B. E., Short, A. T., Smale, A. P., Smith, I. A., Stella, L.,
Tagliaferri, G., Takahashi, T., Tashiro, M., Townsley,
L. K., Tueller, J., Turner, M. J. L., Vietri, M., Voges,
W., Ward, M. J., Willingale, R., Zerbi, F. M., & Zhang,
W. W. 2004, ApJ, 611, 1005
Greif, T. H., & Bromm, V. 2006, MNRAS, 373, 128
Greif, T. H., Johnson, J. L., Bromm, V., & Klessen, R. S.
2007, ArXiv e-prints, 705
Haiman, Z., Abel, T., & Rees, M. J. 2000, ApJ, 534, 11
Haislip, J. B., Nysewander, M. C., Reichart, D. E.,
Levan, A., Tanvir, N., Cenko, S. B., Fox, D. B., Price,
P. A., Castro-Tirado, A. J., Gorosabel, J., Evans,
C. R., Figueredo, E., MacLeod, C. L., Kirschbrown,
J. R., Jelinek, M., Guziy, S., Postigo, A. D. U.,
Cypriano, E. S., Lacluyze, A., Graham, J., Priddey,
R., Chapman, R., Rhoads, J., Fruchter, A. S., Lamb,
D. Q., Kouveliotou, C., Wijers, R. A. M. J., Bayliss,
M. B., Schmidt, B. P., Soderberg, A. M., Kulkarni, S. R.,
Harrison, F. A., Moon, D. S., Gal-Yam, A., Kasliwal,
M. M., Hudec, R., Vitek, S., Kubanek, P., Crain, J. A.,
Foster, A. C., Clemens, J. C., Bartelme, J. W., Canterna,
R., Hartmann, D. H., Henden, A. A., Klose, S., Park,
H.-S., Williams, G. G., Rol, E., O’Brien, P., Bersier,
D., Prada, F., Pizarro, S., Maturana, D., Ugarte, P.,
Alvarez, A., Fernandez, A. J. M., Jarvis, M. J., Moles,
M., Alfaro, E., Ivarsen, K. M., Kumar, N. D., Mack,
C. E., Zdarowicz, C. M., Gehrels, N., Barthelmy, S., &
Burrows, D. N. 2006, Nature, 440, 181
Heger, A., Fryer, C. L., Woosley, S. E., Langer, N., &
Hartmann, D. H. 2003, ApJ, 591, 288
Heger, A., & Woosley, S. E. 2002, ApJ, 567, 532
Helmi, A., Irwin, M. J., Tolstoy, E., Battaglia, G., Hill,
V., Jablonka, P., Venn, K., Shetrone, M., Letarte, B.,
Arimoto, N., Abel, T., Francois, P., Kaufer, A., Primas,
F., Sadakane, K., & Szeifert, T. 2006, ApJ, 651, L121
Iye, M., Ota, K., Kashikawa, N., Furusawa, H.,
Hashimoto, T., Hattori, T., Matsuda, Y., Morokuma, T.,
Ouchi, M., & Shimasaku, K. 2006, Nature, 443, 186
Johnson, J. L., & Bromm, V. 2006, MNRAS, 366, 247
—. 2007, MNRAS, 374, 1557
Johnson, J. L., Greif, T. H., & Bromm, V. 2007, ApJ,
665, 85
Kashlinsky, A., Arendt, R. G., Mather, J., & Moseley,
S. H. 2005, Nature, 438, 45
Kitayama, T., & Yoshida, N. 2005, ApJ, 630, 675
Kitayama, T., Yoshida, N., Susa, H., & Umemura, M.
2004, ApJ, 613, 631
Kneib, J.-P., Ellis, R. S., Santos, M. R., & Richard, J.
2004, ApJ, 607, 697
Lamb, D. Q., & Reichart, D. E. 2000, ApJ, 536, 1
Li, Y., Hernquist, L., Robertson, B., Cox, T. J., Hopkins,
P. F., Springel, V., Gao, L., Di Matteo, T., Zentner, A. R.,
Jenkins, A., & Yoshida, N. 2007, ApJ, 665, 187
Lodato, G., & Natarajan, P. 2006, MNRAS, 371, 1813
Machida, M. N., Tomisaka, K., Nakamura, F., &
Fujimoto, M. Y. 2005, ApJ, 622, 39
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.

Page 12

60. Mackey, J., Bromm, V., & Hernquist, L. 2003, ApJ, 586,
1
Madau, P., Ferrara, A., & Rees, M. J. 2001, ApJ, 555, 92
Magliocchetti, M., Salvaterra, R., & Ferrara, A. 2003,
MNRAS, 342, L25
Mesinger, A., Bryan, G. L., & Haiman, Z. 2006, ApJ,
648, 835
Miralda-Escudé, J. 2003, Science, 300, 1904
Mobasher, B., Dickinson, M., Ferguson, H. C.,
Giavalisco, M., Wiklind, T., Stark, D., Ellis, R. S., Fall,
S. M., Grogin, N. A., Moustakas, L. A., Panagia, N.,
Sosey, M., Stiavelli, M., Bergeron, E., Casertano, S.,
Ingraham, P., Koekemoer, A., Labbé, I., Livio, M.,
Rodgers, B., Scarlata, C., Vernet, J., Renzini, A., Rosati,
P., Kuntschner, H., Kümmel, M., Walsh, J. R., Chary,
R., Eisenhardt, P., Pirzkal, N., & Stern, D. 2005, ApJ,
635, 832
Mori, M., Ferrara, A., & Madau, P. 2002, ApJ, 571, 40
Nagakura, T., & Omukai, K. 2005, MNRAS, 364, 1378
Natarajan, P., Albanna, B., Hjorth, J., Ramirez-Ruiz, E.,
Tanvir, N., & Wijers, R. 2005, MNRAS, 364, L8
Norman, M. L., O’Shea, B. W., & Paschos, P. 2004, ApJ,
601, L115
Oh, S. P., & Haiman, Z. 2002, ApJ, 569, 558
Omukai, K., & Palla, F. 2003, ApJ, 589, 677
O’Shea, B. W., & Norman, M. L. 2007, ApJ, 654, 66
Ostriker, J. P., & McKee, C. F. 1988, Reviews of Modern
Physics, 60, 1
Pelupessy, F. I., Di Matteo, T., & Ciardi, B. 2007, ApJ,
665, 107
Ricotti, M., Gnedin, N. Y., & Shull, J. M. 2001, ApJ,
560, 580
Ricotti, M., Haehnelt, M. G., Pettini, M., & Rees, M. J.
2004, MNRAS, 352, L21
Salvaterra, R., Ferrara, A., & Schneider, R. 2004, New
Astronomy, 10, 113
Santoro, F., & Shull, J. M. 2006, ApJ, 643, 26
Santos, M. R., Bromm, V., & Kamionkowski, M. 2002,
MNRAS, 336, 1082
Santos, M. R., Ellis, R. S., Kneib, J.-P., Richard, J., &
Kuijken, K. 2004, ApJ, 606, 683
Schaerer, D. 2002, A&A, 382, 28
Schneider, R., Omukai, K., Inoue, A. K., & Ferrara, A.
2006, MNRAS, 369, 1437
Shapiro, P. R., Iliev, I. T., & Raga, A. C. 2004, MNRAS,
348, 753
Shapiro, P. R., & Kang, H. 1987, ApJ, 318, 32
Shchekinov, Y. A., & Vasiliev, E. O. 2006, MNRAS,
368, 454
Silk, J., & Rees, M. J. 1998, A&A, 331, L1
Songaila, A. 2001, ApJ, 561, L153
Songaila, A., & Cowie, L. L. 1996, AJ, 112, 335
Spergel, D. N., Bean, R., Doré, O., Nolta, M. R.,
Bennett, C. L., Dunkley, J., Hinshaw, G., Jarosik, N.,
Komatsu, E., Page, L., Peiris, H. V., Verde, L., Halpern,
M., Hill, R. S., Kogut, A., Limon, M., Meyer, S. S.,
Odegard, N., Tucker, G. S., Weiland, J. L., Wollack, E.,
& Wright, E. L. 2007, ApJS, 170, 377
Stacy, A., & Bromm, V. 2007, ArXiv e-prints, 705
Stanek, K. Z., Matheson, T., Garnavich, P. M., Martini,
P., Berlind, P., Caldwell, N., Challis, P., Brown, W. R.,
Schild, R., Krisciunas, K., Calkins, M. L., Lee, J. C.,
Hathi, N., Jansen, R. A., Windhorst, R., Echevarria, L.,
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
79.
80.
81.
82.
83.
84.
85.
86.
87.
88.
89.
90.
91.
Eisenstein, D. J., Pindor, B., Olszewski, E. W., Harding,
P., Holland, S. T., & Bersier, D. 2003, ApJ, 591, L17
Stanway, E. R., Glazebrook, K., Bunker, A. J., Abraham,
R. G., Hook, I., Rhoads, J., McCarthy, P. J., Boyle, B.,
Colless, M., Crampton, D., Couch, W., Jørgensen, I.,
Malhotra, S., Murowinski, R., Roth, K., Savaglio, S., &
Tsvetanov, Z. 2004, ApJ, 604, L13
Susa, H., & Umemura, M. 2006, ApJ, 645, L93
Tominaga, N., Umeda, H., & Nomoto, K. 2007, ApJ,
660, 516
Totani, T. 1997, ApJ, 486, L71+
Tremaine, S., Gebhardt, K., Bender, R., Bower, G.,
Dressler, A., Faber, S. M., Filippenko, A. V., Green, R.,
Grillmair, C., Ho, L. C., Kormendy, J., Lauer, T. R.,
Magorrian, J., Pinkney, J., & Richstone, D. 2002, ApJ,
574, 740
Umeda, H., & Nomoto, K. 2002, ApJ, 565, 385
Whalen, D., Abel, T., & Norman, M. L. 2004, ApJ, 610,
14
Whalen, D., O’Shea, B. W., Smidt, J., & Norman, M. L.
2007, ArXiv e-prints, 708
100. Wijers, R. A. M. J., Bloom, J. S., Bagla, J. S., &
Natarajan, P. 1998, MNRAS, 294, L13
101. Wise, J. H., & Abel, T. 2007, ApJ, 665, 899
102. Wyithe, J. S. B., & Loeb, A. 2003, ApJ, 586, 693
103. Yan, H., & Windhorst, R. A. 2004, ApJ, 612, L93
104. Yoshida, N., Bromm, V., & Hernquist, L. 2004, ApJ,
605, 579
105. Yoshida, N., Oh, S. P., Kitayama, T., & Hernquist, L.
2007, ApJ, 663, 687
106. Yoshida, N., Omukai, K., & Hernquist, L. 2007, ArXiv
e-prints, 706
107. Yoshida, N., Omukai, K., Hernquist, L., & Abel, T.
2006, ApJ, 652, 6
92.
93.
94.
95.
96.
97.
98.
99.