1. A modest overview of HIGH-REDSHIFT DUST Simona Gallerani
(Astronomical Observatory of Rome)
in collaboration with:
R. Maiolino; Y. Juarez; T. Nagao; L. Jiang; X. Fan; C. Willott; R. Mortlock

2. The Dark Ages VIrtual Department
DAVID
The Dark Ages VIrtual Department
S. Bianchi
INAF/Arcetri
B. Ciardi
MPA
P. Dayal
SISSA
C. Evoli
SISSA
A. Ferrara
SNS Pisa
S. Gallerani
OARoma
F. Iocco
IAP
F. Kitaura
SISSA
M. Mapelli
ETH
A. Maselli
MPA
R. Salvaterra
INAF/Milano
S. Salvadori
SISSA
R. Schneider
INAF/Arcetri
L. Tornatore
INAF/Trieste
M. Valdes
IPMU
R. Valiante
Univ. Firenze 2

3. Dust at z~6 why do we care? Early star formation Far-Infrared emission
Trigger molecular hydrogen formation
Allows low mass star formation at low metallicities
Far-Infrared emission
Detectability at mm-submm wavelengths
UV-Optical extinction and reddening
Detectability and interpretation of near-IR light

4. REST FRAME FAR-IR OBSERVATIONS OF z~6 QSOS
Evidence for
large dust masses!
Mdust ~ 108 M
WARM (T~ K) DUST EMISSION
Mdust = 4 ×108 M
Sharc II
SCUBA
MAMBO
Spitzer
F (mJy)
VLA
lrest (mm)
(e.g. Beelen et al. 06; Priddey et al. 03; Bertoldi et al. 03)
In a sample of 33 QSOs at z~6, 30% of these objects have been
detected in the FIR, as in lower redshift samples.
(Wang et al. 2008)

5. REST FRAME NEAR-IR OBSERVATIONS OF z~6 QSOS
Spitzer OBS=24m RF~3m rest
z ~ 6
low-z QSOs
HOT (T~ K) DUST EMISSION
Most z~6 QSOs
have similar hot dust emission
as low-z QSOs
Emission from the hot dust on the
inner side of circumnuclear clouds
not sensitive to the total dust content
...as long as you have some dust!
Jiang et al. 2006

6. REST FRAME NEAR-IR OBSERVATIONS OF z~6 QSOS
But 2 QSOs do NOT show any
hot dust emission!
Never observed in QSOs at low-z.
AGN
Possible cases of very young
objects where dust had no
time to form yet?
Jiang et al. 2006

7. Sources of Dust Mostly produced by evolved low mass stars (AGBs) AGBs
108
AGBs
106
Dust Mass (M)
but bulk of dust (>90%)
produced after 1 Gyr
104
first AGBs
at ~40 Myr...
(e.g. Morgan & Edmunds 2003)
102
0.01
0.1 1 10
Age (Gyr)

8. Sources of Dust at z~6 Age of the Universe < 1 Gyr SN dust ?10 6 1
redshift
108
SN dust ?
QSO dust ?
... ?
AGBs
106
Dust Mass (M)
According to this scenario little dust is expected to be present at z>5 when the Universe was younger than 1 Gyr. Thus, in order to explain the huge amounts of dust inferred from observations different sources of dust have been invoked, as SN or QSO winds.
Is there dust? Evolution of dust mass? Evolution of dust properties? Clues on the dust production mechanism? Account for dust mass? Implications for detectability and observational properties?
104
102
0.01
0.1 1 10
Age (Gyr)

9. SNe as dust factories at z~6
Observational constraints
Theoretical predictions
Ydust = M/SN
(Krause et al. 04; Meikle et 06;
Sugerman et al. 06; Kotak et al. 09)
Ydust = M/SN
(e.g. Todini & Ferrara 01;
Nozawa et al. 03)
Ydust = M/SN
(CasA by Dunne et al. 03)
reverse shocks may also destroy a significant amount of dust
(Bianchi & Schneider 07; Hirashita et al. 08; Nozawa et al. 09)
only 2-20 %
of the dust survive!
(Bianchi & Schneider 07;
Hirashita et al. 08; Nozawa et al. 09)
The presence of 108 M of dust can be explained through
SN production assumin an extreme 100% efficiency of dust
formation or a Top Heavy IMF.
(Dwek et al. 08)

10. Can SNe alone account for the observed dust masses at z~6?
mISM
ISM mass
cleared of dust
Ydust=1.2 M
no destruction
Ydust=0.4 M
Dust Yield per SN (M)
J1148
z = 6.4
tg~400Myr
Zd~0.07
Zd=Md/Mg
Md=2×108M ; Mg~3×1010M
(Dwek et al. 07)
(Bianchi & Schneider 07;
Hirashita et al. 08; Nozawa et al. 09)
The presence of 108 M of dust can be explained through
SN production assumin an extreme 100% efficiency of dust
formation or a Top Heavy IMF.
(Dwek et al. 08)

11. Can SNe alone account for the observed dust masses at z~6?
Dust Yield per SN (M)
J1148
z = 6.4
tg~400Myr
Zd~0.07
M0=5×1010M
Mstar=2×1010 M<<5×1012M
(MBH=5×109M ; Willott 03; Barth 03)
SFR=50 M/yr <<3000 M/yr
(Bertoldi et al. 03)
Zd=Md/Mg
Ydust are overestimated
by at least a factor of 10!
YES!
SNe can make it
(Bianchi & Schneider 07;
Hirashita et al. 08; Nozawa et al. 09)
The presence of 108 M of dust can be explained through
SN production assumin an extreme 100% efficiency of dust
formation or a Top Heavy IMF.
(Dwek et al. 08)

12. Can SNe alone account for the observed dust masses at z~6?
Valiante et al. 09
total dust mass
the AGB
role revisited
carbon dust
other dust
silicate dust
80% AGB-dust
Star formation history by Li et. al. 07; Larson IMF;
Ferrarotti & Gail 06 AGB-dust yields; Bianchi & Schneider 07 SN-dust yields

13. EXTINCTION CURVE at z<4 (Quasars)
dusty wind
Reddening
BAL are the most reddened
sub-class of (type 1) QSOs
SMC-like extinction curve
as for noBAL QSOs at 0<z<2.2
Reichard et al. 2003
Richards et al. 2003; Hopkins et al. 2004

14. EXTINCTION CURVE at z~6 (Quasars)
SDSS
BAL at z=6.23
(Todini & Ferrara 01)
Bluer than any BAL at low z
(Maiolino et al.04)
(Maiolino et al. 04)
Extinction curve consistent with properties of dust theoretically expected from SNe
Also in agreement with Bianchi & Schneider 07 and Hirashita et al predictions

15. EXTINCTION CURVE at z~6 (Gamma Ray Bursts)
Galactic (A3000Å=1.49)
SMC (A3000Å=0.8) K Z J H
GRB050904
at z=6.3
0.5 days
after the burst
SN dust (A3000Å=2)
=QSO J1148 z=6.2
Calzetti (A3000Å=2)
Simple screen case, known intrinsic spectrum
Data from
Haislip et al. 06
Stratta et al. 06

16. A SAMPLE of 33 OPTICAL/NIR SPECTRA of QSOs a 4<z<6.4
(Juarez et al. 2009; Jiang et al. 2007; Willott et al. 2009; Mortlock et al. 2008)
Ftempl template by Reichard et al. 2003, having α=-1.62
A3000 absolute extinction at 3000 Å
A extinction curve normalized at 3000 Å
9 QSOs require substantial (A3000~[ ]) reddening

17. EXTINCTION CURVES Z=0 Z=Z before rs after rs nH=0.1cm-3 mixed
Todini & Ferrara 01
Bianchi & Schneider 07
nH=0.1cm-3
mixed
after rs
nH=1cm-3
unmixed
before rs
20 M
170 M
mixed
nH=1cm-3
unmixed
Hirashita et al. 08
Hirashita et al. 08

18. DUST REDDENING IN 4<z<6.4 QSOs
RF [Å]
RF [Å]

19. DUST REDDENING IN 4<z<6.4 QSOs
RF [Å]
RF [Å]

20. grain size distribution?
DEVIATION FROM THE SMC in z>4 QSOs
Extinction curve
at z<4
Mean Extinction curve
at z>4
different
chemical composition?
grain size distribution?
Evidence for different dust properties in the early Universe

21. BAL vs noBAL REDDENING in z>4 QSOs

22. ON EARLY STAR FORMATION
EFFECTS OF DUST
ON EARLY STAR FORMATION

23. H2 formation on dust grains
(Omukai & Nishi 02; Glover 03; Cazaux & Spaans 04)
Md/Mg = 5×10-4
Md/Mg= 5×10-5
Md/Mg =0
SN M; 0.1 ≤ Yd ≤ 0.3; 50 SNe
(Mvir, zvir) = (109 M, 10)
(Todini & Ferrara 01)
PISN 195 M; Yd ~ 0.3; 1 PISN
(Hirashita & Ferrara 02)
(Schneider et al. 04)

24. Transition from massive PopIII to normal PopII/I stars
Present day
star forming
regions
Salpeter IMF
logξ
2 –
0 –
-2 –
-4 –
-6 -
logM/M -
~1M
>100M
Primordial
star forming
Regions
(Omukai 2000; Abel et al. 2002)
PopIII
PopII
Z>Zcr
(Schneider et al. 2006)
M(frag) [M]
10-6 Z ≤Zcr≤ 5×10-3 Z
(Schneider et al. 2002; 2003; Omukai et al. 2005)

25. Summary Indications of a decreasing amount of dust at z~6
Clear evidence for dust at z~6 and beyond
Indications of a decreasing amount of dust at z~6
(to be verified with ALMA, Herschel and JWST)
SN dust can account for dust masses at z~6, but also the AGB contribution should not be neglected.
Deviation of the mean extinction curve at z>4 from the SMC which characterizes the reddening of QSOs spectra at lower redshifts.
Important implications of dust for the early star formation
Possibly, both AGBs SNe and QSO winds may contribute to the dust enrichment in the early Universe.