Dust and Molecules in Early Galaxies:

1
Dust and Molecules in Early Galaxies Prediction
and Strategy for Observations
Tsutomu T. TAKEUCHI Laboratoire dAstrophysique
de Marseille, FRANCE
Contents
Part I General Introduction Part II Dust
Emission from Forming Galaxies Part III IR
Absorption Line Measurement of H2 in Early
Galaxies Summary
2
Part I General Introduction
3
Dust and Molecules in Early Galaxy Evolution
1. Absorption and Re-Emission of Radiation by Dust
Short wavelength photons are scattered and
absorbed by dust grains and re-emitted as FIR
radiation.
4
2. Dust as a Catalyst of H2 Formation
Dust works as a catalyst for the formation of
molecular hydrogen (H2).
Molecular formation is closely related to the
star formation activity. Especially in an early
phase of galaxy evolution, H2 molecules are very
important coolant of gas to contract to form
stars.
Dust controls the early stage of star formation
history in galaxies! (Hirashita et al. 2002
Hirashita Ferrara 2002).
Without dust, star formation does not proceed
effectively.
5
Motivation
For understanding the physics of galaxy formation
and early evolution, and testing various models,
it is crucial to measure their physical
quantities related to the metal enrichment, dust
production, and molecular gas amount.
We focus on the two systems
1. Young systems with active dust production
Observation of the continuum radiation from dust.
2. Dense systems with little metal/dust
Measurement of H2 through IR absorption lines.
6
Part II Dust Emission from Forming Galaxies
7
Dust Emission Model of Forming Galaxies
1. Model for dust production and radiation
1.1 Dust supply in young galaxies
Dust formation in low-mass evolved stars (RGBs,
AGBs, and SNe Ia) is negligible in a galaxy we
consider here, because of the short timescale
(age lt 108 yr). Dust destruction can also be
negligible by the same reason.
We can safely assume that only SNe II contribute
to the dust supply in young, forming galaxies.
We solve the star formation, evolution of the
strength of UV radiation field, metal enrichment,
and dust production in a self-consistent manner.
8
1.2 Basic framework of our dust emission model
Galaxy formation and evolution

1. Dust supply Nozawa et al. (2003)
2. Star formation history one-zone closed-box model
   with the Salpeter IMF (0.1 lt M lt 100 Msun).
3. Supernova rate calculated from star formation
   rate

Physics of dust grains

1. Optical properties of grains Mie theory
2. Specific heat of grains multidimensional Debye
   model
3. Radiative processes, especially stochastic
   heating of small grains are properly considered
   and included

9
A little more about Nozawa et al. (2003)
Nozawa et al. (2003) proposed a theoretical model
of dust formation by SNe II whose progenitors are
initially metal-free. Two extreme scenarios are
considered for the internal structure of the
helium core of the SN progenitor.

• Unmixed case the original onion-like structure
  of the elements is preserved.
• Mixed case all the elements are uniformly mixed
  in the He core.

We show the results for both cases in the
following.
10
Construction of the SED

• Spherical SF region with radius rSF surrounded
  by dust
• Radiation field strength is calculated from LOB
  and rSF
• LOB evolves according to the SF history.

Based on the SFR and dust size spectrum, the
total SED is constructed by a superposition of
the radiation from each grain species.
Considering the self-absorption by dust, the
final SED is obtained as
11
2. Results
Infrared SED (rSF30pc, SFR1.0Msunyr-1)
12
Infrared SED (rSF100pc, SFR1.0Msunyr-1)
13
The extinction curve of forming galaxies
Age t 107yr.
We will use these extinction curves also in Part
III.
14
3. Observational Implications
3.1 A Local Young Galaxy SBS 0335-052
Blue compact dwarf (BCD) Distance 54 Mpc Very
metal-poor Z 1/41 Zsun Very active star
formation
SFR 1.7 Msun yr-1 (Hunt et al. 2001) Very
young stellar population lt 5 Myr (Vanzi et al.
2000) Very hot dust Tdust gt 80 K Very strong
extinction AV 12-30 mag
15
Comparison of the models with the observed SED
The observed dust SED is roughly reproduced by
the model of unmixed case.
16
3.2 Quest for Forming Dusty Galaxies
Typical physical parameters for high-z small
galaxies
If gas collapses on the free-fall timescale with
a SF efficiency eSF (we assume eSF0.1), SFR of
a galaxy is basically evaluated as follows
(Hirashita Hunt 2004)
If we consider a small clump with gas mass of 108
Msun and adopt rSF 30pc and 100pc, we have a
typical SFR of 10 Msunyr-1. We use these values
for the estimation of dust emission from a
genuine young galaxy.
17
Expected flux for a forming subgalactic clump at
high-z
18
Natural huge telescope gravitational lensing
If we consider a strong gravitational lensing by
a cluster at zlens0.10.2 with dynamical mass of
51014 Msun, it becomes feasible to detect such
galaxies (magnification factor 3040). We can
expect 15 events for each cluster at these
redshifts.
19
Expected flux for a forming subgalactic clump at
high-z II
20
Part III IR Absorption Line Measurement of H2 in
the Early Galaxies
21
IR Absorption Line Measurement of H2 in the Early
Galaxies
1. Basic Idea
H2 molecules the predominant constituent of
dense gas.
Local Universe Molecules containing heavy
elements (e.g., CO, etc) are good tracers of the
amount of H2.
High-z systems in their first star formation They
are very metal-poor, and we need a special
technique for measuring the amount H2 directly..
22
Petitjean et al. (2000) and subsequent studies
showed a direct measurement of H2 in UV
absorption lines. Their target is damped Lyman-a
absorbers (DLAs).
Transition probability A of ionizing/dissociation
lines is so large that they are useful for
detecting thin layers and small amounts of the
molecular gas, but not useful for detecting dense
gas clouds, as those of our interest.
Then, H2 has well-known vibrational and
rotational transitions in the IR. Their
transition probabilities are very small because
the H2 is a diatomic molecule of two identical
nuclei, and has no allowed dipole transitions.
The vib-rotaional and rotational line emission of
H2 are useful for analyzing dense (n gt 10 cm-3)
and hot (T gt 300 K) gas.
23
Unfortunately, direct measurement of the H2
emission lines is very difficult for distant
galaxies (Ciardi Ferrara 2001).
If, however, there is a strong IR continuum
source behind or in the molecular gas cloud,
absorption measurements of these transition lines
can be possible (Shibai, Takeuchi, Rengarajan,
Hirashita 2001, PASJ, 53, 589)!
Such observation will be feasible by the advent
of the proposed space missions for large IR
telescope, like SPICA, etc.
24
SPICA One of the Observational Possibilities
SPICA (Space Infrared Telescope for Cosmology and
Astrophysics) is the next-generation IR mission,
which is to be launched by the Japanese HIIA
rocket into the L2 point. This mission is
optimized for M- and FIR astronomy with a large
(3.5 m), cooled (4.5 K) telescope. The target
year of launch is 2010.
http//www.ir.isas.jaxa.jp/SPICA/index.html
25
2. Calculation
Assume a uniform cool gas cloud with kTex ltlt hn,
then the optical thickness of the line absorption
is expressed as
where u and l upper and lower states, gu and gl
degeneracy of each state, Aul Einsteins
coefficient, Nl column density of the molecules
in the lower state, and DV line width in units
of velocity.
26
Assumption almost all the molecules occupy the
lowest energy state.
Absorption line flux in the extinction free case
is
where Dn is a line width in units of frequency,
S continuum flux density of the IR source behind
the cloud. Subscript 0 means extinction-free.
We consider DV100 kms-1 for the line, and
S10mJy as a baseline model. If the line optical
thickness is smaller than 0.01, it is very
difficult to detect. We therefore assume
tline,00.01.
27
The dust extinction is introduced as
where Al/AV extinction curve, and AV/NH
normalization. We use the Galactic extinction
(Mathis 1990) as a baseline model. We also see
the effect of different extinction curves. The
extinction scales with metallicity Z. Using this,
absorption line flux with extinction is obtained
as
28
Summary of the parameters for this calculation
The detection limits is for SPICA (Ueno et al.
2000).
29
Considered hydrogen lines in the IR
(Shibai et al. 2001, PASJ, 53, 589)
30
3. Results
3.1 Absorption lines vs. dust extinction
(metallicity)
If the metallicity is one solar, we cannot detect
these lines because of strong extinction. But if
Z 0.01 Zsun, they can be detected by SPICA.
31
3.2 Effect of extinction curve
Solid line is the result for the Galactic
extinction, while the dotted line is the
mixed-case extinction curve, and dashed line is
for the unmixed-case one. The result is sensitive
to the extinction curve when the column density
of the gas is high.
32
3.3 Possible background source
We consider QSOs, especially lensed ones. If we
put some known QSOs at z5, they have flux
densities around 10mJy.
Considering a 60K blackbody, 17, 28, and 112 mm
lines will be suitable for this observation, but
2 mm line is hard to detect because of the weak
continuum.
33
4. What can we learn?
Consider a protogalactic cloud of M 1011 Msun.
Since the radius R is a few kpc in this case, we
have its column density
where f gas mass fraction of molecular clouds.
This fraction can be very high ( 1) when NH is
high enough (Hirashita Ferrara 2005).
Its evolution occurs in a free-fall timescale,
much shorter than the cosmic evolution timescale,
e.g., Hubble time.
Observed properties are specific to the redshift
at which the cloud absorption is measured.
34
We obtain z and DV (velocity dispersion) of
primordial gas clouds from this observation.
These quantities tell us their dynamical
evolution through the structure formation theory.
Collapse of a massive cloud (M 1011Msun) at z
lt5
Basically observed in the IR, SPICA will be useful
Population III objects (M 106-9Msun) at z gt 5
Observed in the submm, ALMA will be required
35
Summary
36
1. Summary of dust emission model

1. We constructed a model for the SED of forming
   galaxies based on a new theory of dust production
   by SN II.
2. The model (unmixed case) roughly reproduced the
   observed SED of a local low-metallicity dwarf
   SBS0335-052, which has a peculiar strong and
   MIR-bright dust emission.
3. We also calculated the SED of a very high-z
   forming small galaxy. Although it may be
   intrinsically too faint to be detected even by
   ALMA, gravitational lensing can make it possible.

37
2. Summary of IR Absorption Measurement of H2

1. We proposed a method to measure the amount of H2
   in primordial low-metallicity cloud in absorption
   in an IR spectra of QSOs.
2. If the metallicity of the cloud is low (Z 0.01
   Zsun), dust extinction is expected to be so weak
   that 17 and 28mm lines are detectable by SPICA
   for objects at z lt 5. Small very high-z
   population III objects will be detected by ALMA.
3. By this method, we can trace back the dynamical
   evolution of early collapsing objects at very
   high redshifts.

38
(No Transcript)
39
Dust grain species produced by SN II
40
Grain size spectrum of dust produced by SN II
(Nozawa et al. 2003, ApJ, 598, 785)
41
Chemical evolution (a little more)
Closed-box model is assumed.
Time evolution of the mass of ISM
where SFH is assumed to be constant, and we
adopted Salpeter IMF
Remnant mass (fitting formula)
42
Important transitions of H2 molecules
43
Equivalent width of some IR lines
The second line follows by the optically thin
condition.
44
SPICA sensitivity
45
H2 17mm line for various Z