Observations of molecules in high redshift galaxies

Observations of molecules in
high redshift galaxies
Kirsten K. Knudsen
Chalmers University of Technology
Onsala Space Observatory
Outline: Small molecules; Large molecules
Excitation ; In the ALMA perspective

IN A GALAXY FAR AWAY ...
Looking back in time!
Redshift
z
Universe
age / 10 9 yr
0 13.7
1 5.9
2 3.3
4 1.6
6 0.95
8 0.65
1000 0.0004
Evolution of galaxies: production of stars, metals, and molecules
Introduction 2

CO AT HIGH REDSHIFT
Only
high redshift
Introduction
→ MH2
molecular
gas mass
Both local and
high redshift
→ star formation rate
3From: a.o. Knudsen et al., 2009, 2011

QSOS: FEW MOLECULAR LINE OBSERVATIONS
• For z>2:
~35000 quasars known
~30 have CO detections
100000
10000
1000
100
All quasars
QSOs w/ CO obs
• Small number due to:
sensitivity, bandwidth, ...
• Typically single object studies
10
1
2 3 4 5 6
redshift
• ~ 1/3 QSOs are FIR bright 4
CO in quasars
redshift z 0 2 4 6
Universe age 13.7 3.3 1.6 0.95

CO IN z = 4 - 5 QUASARS
CO in z=4-5 quasars
Knudsen, Bertoldi et al.
5

REDSHIFT
Example: J0808
zsdss=4.4562,
zCO=4.4173
Vdiff = -2145 km/s
SDSS Plate 1780, MJD 53090, Rerun 26, Fiber
529
Some other cases have
large differences in
literature redshifts with
Vdiff up to 7000 km/s
CO in z=4-5 quasars
See the CAS Explorer page for more information on this object
Knudsen, Bertoldi et al.
6

248 GONZALEZ ET AL.
H2 AT Z=2.8
No. 1, 2010
|magnification| ~ 80 x
A z = 2.79 LENSED LIRG BEHIND THE BULLET CLUSTER
Figure 3. HST imaging of a 14 ′′ × 9 ′′ region encompassing the lensed galaxy. The image on the left is a color composite using the F850LP, F110W
bands for blue, green, and red, respectively. Despite the bright foreground galaxy, the lensed galaxy can be seen as a red arc with two lobes. On the right,
Table 1
F160W image of the same region after the bright foreground objects have been modeled with GALFIT and subtracted. In this image, we also mask the
three brightest objects for clarity. The arc, which is clearly visible in this image, breaks up into two lobes corresponding
Observed
to images
Fluxes
A and
and
B
Magnitudes
in the IRAC a
i
Paper I. Crosses denote the positions previously reported based upon IRAC photometry. The offsets are consistent with the uncertainties in the IRAC po
source blending. North is up and east is to the left.
Quantity
V
Lensed dwarf galaxy
(A color version of this figure is available in the online journal.)
f (6.2 µm) 1.4 ± 0.2 × 10
M*
f (7.7 ~ µm) 4x10 9 Msun
6.3 ± 1.2 × 10
the spectroscopic redshift and improved astrometric positions. BintheF160W
f (8.6 µm)
observations is consistent
5.2
with
± 3.3
the
× 10
2
The former has only a modest impact, whereas the latter have ratio. Because f (H 2 S(4)) we are probing near the critical 5.8 ± curves 1.9 × 10 w
amoresignificantimpactbecauseofthesteepmagnification are strong f (H 2 S(5)) magnification gradients, this 2.5 should ± 1.2 not × 10 n
H2:
gradient near the critical curve.
be thef case (7.7 Tex µm)/f(6.2 = 377 µm) K
unless the stellar emission and PAH emi 4.5
In the current discussion, we are concerned primarily with similar m F spatial 160W distributions within the galaxy. Thus, 23.80 ±
MH2 = 2x10 8 Msun
images A and B. The WFC3 astrometry moves both images A data further support the picture that in this galaxy the
and B slightly closer to the critical curve and consequently far-infrared Note. a emission All quotedis values dominated are for by thestar combination formation. of
raises the derived magnifications. At the location of image
Awecomputeamagnification|µ A | ∼ 32 and at image B
3.5. Stellar Population Age and Mass
we obtain |µ
use a power law to fit the underlying conti
B | ∼ 69. The measured flux ratio of the two
images from Paper I is 1.47 ± 0.05. To estimate the range the We redshift present infrom Figure the4 two the full strongest spectralPAH energy feat d
of plausible magnification for the blended IRS spectrum, we
for7.7 theµm), combination obtaining of images z = 2.791 A and± B, 0.007. including This all
take individually the predicted magnifications for A and B at
data. theGiven photometric the redshift, redshifts we canin now the repeat literature the analy
Gonzalez et al. 2010 (z ∼
Figure
the
2. IRS
locations
spectrum
of
of
peak
the galaxy,
emission
taken
for
in
each
the long-low
arc and
mode.
combine
The data
these
are SED et al. presented 2008; Gonzalezetal.2009; in Paper I to refine the physical Rexetal.2 con
presented
H2 at high-z
as the shaded histogram, with the uncertainties shown by the hashed
with the observed flux ratio. 7
histogram. The dot-dashed line indicates 8 We find that the total combined
the system.
the best fit from PAHFIT, while the The derived fluxes for the PAH featu
We first use HyperZ (Bolzonella et al. 2000) to

0, L3 (2011)
H2O AT HIGH REDSHIFT
para H2O 202 - 111 987.9GHz
Lensed Herschel source:
SPD.17b, z=2.305
H2O / CO similar to Mrk231
(Omont et al. 2011)
t
.
Fig. 1. Spectrum ofH2O the para H 2 O 2 02 −1 11 emission line towards
616 - 523 22.2GHz maser
SDP.17b, where the velocity scale is centred on its observed frequency
at 298.93 GHz (corresponding to z = 2.3049). The rms noise is
∼4.7 mJy/beam in 31.6 MHz channels. A Gaussian fit to the H 2 Ospectrum
is shown as a solid line while the dotted line shows the underlying
dust continuum emission. The H 2 Olineisclearlyasymmetricandnot
well fit by a Gaussian profile.
the Plateau de Bure Interferometer (PdBI) and future facilities
such as ALMA and NOEMA. Tentative detections of H 2 Owere
H2O at high-z
reported in the Cloverleaf for the 2 20 −1 11 transition (Bradford
Lensed quasar:
MG 0414+0534, z=2.64
Possibly associated with circumnuclear
accretion disk or jet
(Impellizzeri et al. 2008)
8

HIGH DENSITY TRACERS
1034 RIECHERS ET AL. Vol. 725
APM08279 (z=3.9)
L18 GARCÍA-BURILLO ET AL. Vol. 645
Parameter
TABLE 1
Observational Results
We observed APM 082795255 with the IRAM six-element
array in its B configuration on 2006 March 13. The spectral
correlator was adjusted to z p 3.911 and centered on redshifted
HCO (5–4) (rest frequency at 445.903 GHz) to cover an effective
bandwidth of 580 MHz (equivalent to ∼1800 km s 1 ).
The size of the synthesized beam was 1.46#1.20 (P.A. p
94) at the observing frequency (90.797 GHz). APM
082795255 was observed for a total integration time of
9.5 hr on-source. During the observations, the atmospheric
phase stability on the most extended baselines was always better
than
Fig. 1.—Spectrum of the HCO (5–4) line detected toward the peak of
Figure 1. CARMA spectra of the HCN(J = 6→5), HCO + 20, typical for excellent winter conditions. The absolute
flux density
(J = 6→5) (left), HNC(J = 6→5) (middle), and continuum integrated emission line intensity at 2.76, in 2.71, APM and 082795255 2.84 mm at (left (a, d) p to (08
right)
Cloverleaf
toward APM 08279+5255 at a resolution
(z=2.6)
h 31 m 41 ṣ 73,
of ∼85 km s −1 scale was calibrated using 3C 84 and 524517.4). Velocities ( v v
(31.25 MHz). The velocity scales are relative to the tuning frequencies of
0 )havebeenrescaledwithrespecttotheCO
0836710 and should be accurate to better than 10%. We have redshift of z p 3.911 (Downes et al.
108.611
1999). The
GHz
dashed
(left)
line shows
and
the Gaussian
110.751 GHz (middle) and the central frequency used of both theLSBs GILDAS at 105.391 packageGHz for the (right). data reduction The rms and per velocity analysis. bin are fit to 0.95, the HCO 0.88, (5–4) and ∼0.65–0.92 emission, and the mJy highlighted (left tochannels right). in intervals I
The solid curves show simultaneous Gaussian fits to the line and continuum emission where applicable.
and III identify the range for line-free continuum emission.
The average 1j noise channels of 20 MHz width is estimated
(A color version of this figure is available in the online to be journal.) 0.5 mJy beam
HCO 1 . The phase + tracking center of these
observations is at J2000 (a 0 , d 0 ) p (08
1-0
h 31 m 41 ṣ 57, 524517.7), tracted) toward the peak of integrated intensity at (a, d) p
Table 1 coincident with that used by Wagg et al. (2005) in their HCN (08 h 31
Table m 41 ṣ 73, 524517.4) (see Fig. 2). Within the errors, the
2
Measured Line Fluxes and Luminosities in
observations.
APM 08279+5255
To calculate luminosities, we assume
Measured
a L cosmology
described by H 0 p 70 km s 1 , Q L p 0.7, and Q m p vious millimeter continuum maps (Downes et al. 1999; Wagg
peak position is in agreement with that determined from pre-
Continuum Fluxes in APM 08279+5255
Line I 0.3 (Spergel et L al. ′ 2003). Throughout this Letter, ν the velocity et al. 2005). The velocity coverage encompasses the full extent
obs λ obs λ rest S ν
(Jy km s −1 scale is
) (10 10 referred
µ −1
L K to km the CO redshift, s−1 pc 2 z p 3.911.
of the HCO line emission. Data from the line-free sidebands
)
(GHz) (mm) with velocities (µm) below 400 km s(mJy)
1 and above 550 km s 1
HCO + (J = 6→5) 1.01 ± 0.19 3.0 ± 0.6
110.7 2.71(intervals I and 551III, defined in2.17 Fig. ± 1) 0.19 were combined with
3. RESULTS
HCN(J = 6→5) 1.65 ± 0.19 4.9 ± 0.6
108.6 2.76natural weighting 562 to estimate2.08 the ± continuum 0.22 emission at
We have detected the HCO (5–4) line and the
HNC(J = 6→5) 0.86 ± 0.24 2.4 ± 0.7
105.4 a dust continuum
emission (at 670 mm), both emitted in the submilli-
value of 1.2 0.3 mJy at 93.9 GHz from Downes et al. (1999).
90.8 GHz to be ∼1.2 0.13 mJy, in good agreement with the
2.84 579 2.04 ± 0.08
HC 3 N(J = 57→56) a

HIGH DENSITY TRACERS (HCN)
No. 1, 2010 HCN, HNC, AND HCO + (J = 6→5) IN APM 08279+5255 (z = 3.91) 1
(CN) and above (HC 3 N), a more complex scenario would
required to explain the different HCN, HCO + ,andHNC
ratios in the J = 5→4 and6→5 transitions. In this cas
will become necessary to investigate the different efficie
with which the IR radiation field excites their bending mode
more detail, which may lead to high but different excitation
the rotational ladders of these molecules.
APM08279:
Strong IR radiation field.
Collisional excitation not enough,
likely radiative excitation
5.2. HNC/HCN Ratio
HCN ratios and high HNC excitation. Two scenarios w
brought forward to explain these high ratios: IR pumping a
or increased abundances of HNC in the presence of X-
dominated regions (XDRs, often found in regions impacted
emission from active galactic nuclei; Aalto et al. 2007b).
the one hand, we find that HCN(J = 6→5)/HCO + (J = 6→
> 1, which is the opposite to what is expected in the X
scenario (Aalto et al. 2007b). However, we note that our L
models show that the HCN(J = 6→5) line (and thus, lik
also the HNC J = 6→5line)isonlymoderatelyopticallyth
APM 08279+5255 also fits well into the IR-pumping scena
in nearby infrared-luminous galaxies, the gas is warm eno
to efficiently pump HNC (through the 21.6 µm bendingm
with an energy level of hν/k = 669 K and Einstein A coeffic
6(V666/71417)7/10/07 KNUDSEN ETRecent AL. studies of HCN and HNC emission in nearby Vol. infra 66
luminous galaxies have revealed sources with high HN
Riechers et al. 2011
abundance effects thus cannot be ruled out. On the other ha
the HNC(J = 6→5)/HCN(J = 6→5) ratio of 0.50 ± 0.1
Figure 4. HCN excitation ladder (spectral line energy distribution; points) and
LVG models (lines) for APM 08279+5255, accounting for both collisional and
radiative excitation. The HCN(J = 5→4) data point is from Weiß et al. (2007).
The models give n gas = 10 4.2 cm −3 , T kin = T IR = 220 K, and infrared field-filling
factors of IR ff = 0.3–0.7 (shown in steps of 0.1).
(A color version of this figure is available in the online journal.)
of A
NGC253 IR = 5.2 s −1 ), but not HCN (through the 14.0 µm bend
(Knudsen et al. 2007)
Fig. 8.—HCO + (right)andHCN(left) line SEDs toward the nucleus of NGC 253 in a 27 00 beam. The data points are from Nguyen-Q-Rieu al. (1992; triangle
High
is currently
Paglione density ettracers
not possible to properly constrain similar models mode hν/k = 1027 K and A IR = 1.7 s −1 ;Aaltoetal.200
al. (1997) and Paglione (1997) ( HCN and HCO + , respectively, open squares), and this work ( Table 1; solid squares and circle). The lines in10
both diagra
for the HCO + and HNC line excitation. However, due to leading to a high HNC/HCN L ′ ratio in high-J transitions
are selected single-component LVG model fluxes for [n(H ), T ] combinations; solid line: 10 5.2 cm 3 , 50 K; dashed line: 10 4.9 cm 3 ,150K;dottedline:10 5.5 cm

HIGH DENSITY TRACERS (HCN):
SFR
11 11
uminosities L FIR 1 7 # 10 L, ( LIR 1 9 # 10 L,
)
erage ratio of 880 with a 1 j dispersion of (330, HCN
(K km s 1 pc 2 ) 1 .Thesearetheclosestanalogstothe
o of the high-z HCN detections (Cloverleaf and SMM
ad to L /L ′
FIR HCN values within the range expected from
axies, similar to the ratio in the best-known ULIRGindex = 1
′
ith L /L p 1340 L (K km s 1 pc 2 ) 1 FIR HCN ,
.Thethree
tions have ratios about twice that of Arp 220. Only
e 65 local galaxies have ratios in this range (2200–
′
r measurements yield high upper limits to L FIR/L
HCN,
distribution expected from local galaxies including
here are also three weak upper limits that are not very
mbining the detections and upper limits, the star forte
per solar mass of dense gas, measured by
ishigherinEMGsthaninthelocaluniverse,including
y a factor of ∼2.0–2.5.
be noted that with the exception of four SMGs, in-
HCN detection reported here, all other EMGs in our
M dense
st known AGNs. Were the FIR luminosities of these
e corrected by the amount contributed by their AGNs,
plotted would, in principle, come into better agreement
w-z correlation. However, for the three quasars where
ions have been calculated (F102144724, D. Downes
olomon2007,inpreparation;Cloverleaf,Weissetal.
High density tracers
GAO ET AL. Vol. 660
Fig. 4.— LFIR vs. L HCN/LCO. All galaxies with a high dense molecular gas fraction
11
of L HCN/LCO 1 0.06, whether local or high z,areIR-luminous( LFIR 1 10 L,
).The
high-z EMGs are indicated with filled circles, local LIRGs and ULIRGs with crosses,
and normal spirals with open circles. The high-z EMGs show a high ratio of dense
to total molecular gas similar to ULIRGs but even higher FIR luminosity.
e.g. Gao ea. 2007
082795255, Weiss et al. 2005, 2007) the corrections the global HCN emission is enhanced relative to CO in galaxies
11

THE EYELASH - EXAMPLE OF LENSING
CO(1-­‐0) CO(3-­‐2) CO(4-­‐3) CO(5-­‐4)
CO(6-­‐5) CO(7-­‐6) CO(8-­‐7) CO(9-­‐8) CO(10-­‐9)
[CI](1-­‐0) [CI](2-­‐1) HCN(3-­‐2)
CO excitation
The ‘Eyelash’ –
the (apparently) brightest source known
[serendipitous LABOCA detecSon]
lensing gives 200pc resoluSon
in source plane
àvery luminous, compact HII reg.
Danielson et al. 2011
12

THE EYELASH - EXAMPLE OF LENSING
1692 A. L. R. Danielson et al.
LVG model
PDR model
Figure 2. The integrated 12 CO SLED for SMM J2135 showing that the SLED peaks around J upper = 6(similartoM82,butwithproportionallystronger
12 CO(1–0)). The central panel shows the results of LVG modelling applied to the integrated SLED, which requires two temperature phases to yield an adequate
fit. In this model, they have characteristic temperatures of T kin = 25 and 60 K and densities of n = 10 2.7 and 10 3.6 cm −3 ,respectively(weassociatethesetwo
Using this dynamical mass and the minimum gas mass, we place
a lower limit on the gas fraction of M gas /M dyn 0.19. This ratio
is similar to the ratio for typical starburst nuclei and the ratio of
30 per cent found in M 82 (Devereux et al. 1994). Compared to other
high-redshift submillimetre galaxies (SMGs), using the 12 CO(3–2)
observations from Greve et al. (2005) and adopting α = 0.8 and
R 3,1
CO = 0.58 excitation ± 0.05 from Ivison et al. (2010c), we find a median
The ‘Eyelash’ –
two components based on CO
excitation and dust emission:
1) low-dense, ~25K, 1-2kpc
2) high-dense, ~60K, ~200pc
phases with a cool, extended disc and hotter, more compact, clumps). The right-hand panel shows a similar comparison but now using a PDR model (Meijerink
et al. 2007; see Section 3.2.1), which again requires a combination of both low- and high-density phases to adequately fit the integrated SLED.
high cosmic-ray flux (see below), it is probable that momentumdriven
outflows will result. These outflows could be either photondriven
(Thompson, Quataert & Murray 2005) or cosmic ray driven
(Socrates, Davis & Ramirez-Ruiz 2008) and will impact both the
dynamics of the gas in the system and the steady-state assumption
Danielson et al. 2011
in our PDR modelling (Section 3.2.1).
We can now compare the star formation efficiency (SFE) 13 in

CO LINE SEDS
IRAM 30m CO SED survey
(1, 2, 3mm bands) Weiss, Walter, Downes, Henkel, in prep.
CO excitation
14

CO LINE SEDS
• single gas component
• QSO turnover: >6-­‐5
• SMM turnover: 5-­‐4
• compact (

MASSIVE GAS RESERVOIRS
z~2 gal’s: not extreme starbursts, but massive gas reservoirs
CO(2-1)
PdBI
HST
§ 6 of 6 detected in CO, ~10 kpc size
§ M gas > 10 10 M o ~ high-z HyLIRG !
! (SMG, QSO host)
But:
§ SFR < 10% HyLIRG
§ 5 arcmin -2 (vs 0.05 for SMGs)
=> common, ‘normal’ high-z galaxies
- SFR/M * const. w/ M * : ‘pre-downsizing’
Daddi ea. 2007, 2008, 2009, Tacconi et al. 2010, Genzel et al. 2010
Text
16

OTHER GALAXIES, DIFFERENT EXCITATION
I CO [Jy km s -1 ]
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
BzK-21000
BzK-4171
BzK-16000
1 2 3 4
Rotational Quantum Number, J
Fig. 5.— CO excitation ladder or CO SED for BzK-21000 (solid
black circles). The black solid, dashed high zand dotted lines show representative
LVG models 1, 2 and 3 from Dannerbauer et al. (2009),
respectively. The open stars and open circles show the CO 1−0 and
CO 2 − 1integratedfluxdensitiesforBzK-4171andBzK-16000,
scaled to the CO 2 − 1emissionofBzK-21000forcomparison. The
open triangle, open square and asterisk symbols illustrate the CO
SED of the Milky Way, the Antennae overlap region and the spiral
galaxy NGC6946, normalized to the CO 2 − 1emissionofBzK-
21000. The LVG models that reproduce the data for these objects
are shown as gray solid low lines. z A small horizontal shift has been
applied to the data points to enhance the visibility.
Milky Way
Cold molecular gas in massive disk galaxies at z =1.5
Antennae overlap
NGC6946
Inner MW
L FIR /L’ CO
gas consumption lifetimes of ∼ 0.2 −
ble 2). These values are comparable t
Daddi et al. (2010a) and are similar tho
parable galaxies at redshifts ∼ 1−3 (Tac
K-band
§ CO excitation = Milky Way (but M gas > 10x MW)
§ L FIR /L’ CO = MW few x10 8 yrs
3.2. Excitation of the molecu
In this section we study the propert
ularMW
gas in our BzK galaxies. All of
previously been detected in the CO 2 −
However, only BzK-21000 has been de
3 − 2 line (Dannerbauer et al. 2009).
Figure 5 shows the velocity integrated
BzK-21000 as a function of rotational q
J. Also shown are the CO 1 − 0andCO
line
2 arcsec
fluxes of BzK-4171 and BzK-1600
the CO 2 − 1 intensity of BzK-21000
and the normalized CO intensities for
the Milky Way (Fixsen et al. 1999), the
region (Zhu et al. 2003) and the spiral
(Bayet et al. 2006).
From the velocity integrated line flu
the brightness temperature line ratios as
(I 21 )/I 10 )×(ν 10 /ν 21 ) 2 ,whereT 21 and T
ness temperatures, I 21 and I 10 are the
and ν 21 and ν 10 are the observed frequ
2 − 1and1− 0 emission lines, respec
SMMJ163556
z = 4.04
no CO(4-3) - why?
(Dannerbauer we expect et al. 2009; thisAravena ratio toet be al. r2010)
21 =1. In
and 0.28 CO ± excitation 0.08 Jy km s −1 in BzK-4171, BzK-21000 and
measure brightness temperature17line ra
+0.61 +0.23

DISK DIMMING DUE TO CMB
CO excitation
From Axel Weiss
18

CO, C + , [CI] AT HIGH REDSHIFT
C +
[CI]
ALMA
[CII] 2 P 3/2 → 2 P 1/2
[CI] 3 P 2 → 3 P 1
[CI] 3 P 1 → 3 P 0
Band 10 (787-950 GHz)
Band 9 (602-720 GHz)
Band 8 (385-500 GHz)
Band 7 (275-370 GHz)
Band 6 (211-275 GHz)
Band 4 (125-163 GHz)
Band 3 ( 84-119 GHz)
18
16
14
12
CO J upper
10
CO Jupper
8
EVLA
Q 0.7cm (40-50 GHz)
Ka 1.0cm (26-40 GHz)
K 1.3cm (18-26 GHz)
MeerKAT
High (8-14.5GHz)
CO at high redshift
6
4
2
redshift
0 5 10 15 20
Redshift
Courtesy: J. Kurk
19

MASSIVE GALAXIES AT Z=1.5-2.5, ALMA SIMULATIONS
From: Fabian Walter
ALMA simulations

QSOS AT Z=6 WITH ALMA
From: Fabian Walter
ALMA simulations

PAHS IN HIGH-Z GALAXIES
No. 1, 2009
MID-IR SPECTROSCOPY OF SUBMI
MID-IR SPECTROSCOPY OF SUBMILLIMETER GALAXIES 673
the I
1998
mino
indic
can b
log (LFIR)
whe
resp
×10
Alex
ame
inclu
6 8 12 log (L7.7µm)
median-combined rest-frame composite spectrum of 22 SMGs detected in our sample, with the shaded area representing the 1σ sample standard
me optical redshifts (C05), except for the five SMGs in which PAH features suggest Figure an alternate 11. PAH luminosities redshift (see as aSection function4.1). of IR The luminosity composite for thespectrum
SMGs in our
ong PAH emission at 6.2, 7.7, 8.6, 11.3, and 12.7 µm, with an underlying red continuum. sample and All for the those spectra in Pope areet normalized al. (2008). We byalso their include flux at aperture-corrected
λ = 8.5 µm
ee Section 3.1 for details). Right: composite spectra for subsets of SMGs in threePAH separate luminosities redshift of bins. low-redshift Thesenuclear demonstrate starbursts that (Brandl at λ > et10 al. 2006). µm our The
is dominated by sources at z

SUMMARY
• Growing number of detections of molecular gas at high-z!
• Detections of HCN, HCO + , CN, H2O, only in a few extreme objects
(quasars and lensed starburst galaxies).
• Excitation of gas shows a variety of conditions
• Example: The Eyelash - CO line SED shows the a composite of
low-dense and high-dense gas.
• Massive galaxies with low excitation molecular gas.
• At high-z, the CMB will impact the observations of cold, diffuse
gas in e.g. disks.
• High-z PAH studies allowing to probe differences between high-z
starbursts and local ULIRGs.
• With full-array ALMA, astrochemistry in high-z galaxies will
become possible.
Summary
23