Transcript Dust and Molecules in Early Galaxies

Dust and Molecules in Early Galaxies:
Prediction and Strategy for Observations
Tsutomu T. TAKEUCHI
Laboratoire d’Astrophysique 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
Part I
General Introduction
Dust and Molecules in Early Galaxy Evolution
1. Absorption and Re-Emission of Radiation by Dust
Star formation
Heavy element production
Short wavelength photons are
scattered and absorbed by
dust grains and re-emitted as
FIR radiation.
Dust formation
UV
Dust
FIR
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.
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.
Part II
Dust Emission from Forming
Galaxies
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 < 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.
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 < M < 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
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.
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
2. Results
Infrared SED (rSF=30pc, SFR=1.0Msunyr-1)
Infrared SED (rSF=100pc, SFR=1.0Msunyr-1)
The extinction curve of forming galaxies
Age t = 107yr.
We will use these extinction curves also in Part III.
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: < 5 Myr (Vanzi et al. 2000)
Very hot dust: Tdust > 80 K
Very strong extinction: AV = 12-30 mag
Comparison of the models with the observed SED
The observed dust SED is roughly reproduced by the model
of unmixed case.
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 eSF=0.1), 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.
Expected flux for a forming subgalactic clump at high-z
ALMA detection limits (64 antennas, 8 hours).
Herschel confusion limits by Lagache et al. (2003).
Natural huge telescope: gravitational lensing
If we consider a strong gravitational lensing by a cluster
at zlens=0.1–0.2 with dynamical mass of 5×1014 Msun, it
becomes feasible to detect such galaxies (magnification
factor ~ 30–40).
We can expect 1–5 events for each cluster at these
redshifts.
Expected flux for a forming subgalactic clump at high-z II
Part III
IR Absorption Line Measurement
of H2 in the Early Galaxies
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..
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 > 10 cm-3) and hot (T > 300 K) gas.
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.
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
2. Calculation
Assume a uniform cool gas cloud with kTex << 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: Einstein’s coefficient, Nl: column density of the
molecules in the lower state, and DV: line width in units of
velocity.
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 DV=100 kms-1 for the line, and S=10mJy as a
baseline model. If the line optical thickness is smaller than
0.01, it is very difficult to detect. We therefore assume
tline,0=0.01.
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
Summary of the parameters for this calculation
The detection limits is for SPICA (Ueno et al. 2000).
Considered hydrogen lines in the IR
(Shibai et al. 2001, PASJ, 53, 589)
3. Results
3.1 Absorption lines vs. dust extinction (metallicity)
Z = 1 Zsun
Z = 0.01 Zsun
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.
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.
3.3 Possible background source
We consider QSOs, especially
lensed ones. If we put some
known QSOs at z=5, 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.
17, 28mm: IR
112mm: submm
SPICA
ALMA
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.
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 <5:
Basically observed in the IR, SPICA will be useful
Population III objects (M ~ 106-9Msun) at z > 5:
Observed in the submm, ALMA will be required
Summary
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 SBS0335052, 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.
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 < 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.
Dust grain species produced by SN II
Grain size spectrum of dust produced by SN II
(Nozawa et al. 2003, ApJ, 598, 785)
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)
Important transitions of H2 molecules
Equivalent width of some IR lines
The second line follows by the optically thin condition.
SPICA sensitivity
Herschel
SPICA
H2 17mm line for various Z