ZEUS/CSO Observations of Far-Infrared Fine Structure Lines: Diagnostics of Star

Formation in the Early Universe

Gordon Stacey Cornell University

S. Hailey-Dunsheath, T. Nikola, C. Ferkinhoff, S.C. Parshley, D. Benford, J.G. Staguhn, C. E. Tucker, & N. Fiolet

1

Far-IR Fine Structure Lines   Most abundant elements are O, C, N   Species with 1,2,4 or 5 equivalent p electrons will have ground state

terms split into fine-structure levels   For O, C, N they lie in the far-IR where extinction is not an issue

 Collisionally excited / optically thin ⇒ cool the gas – trace its physical conditions

 Reveal the strength and hardness of ambient UV fields.

2p: O++; N+; C0

0

3P2

3P0

3P1

[OIII] 51.8 µm

[OIII] 88.3 µm

442

163

2P3/2

2P1/2 0

1p: O+++; N++; C+

[OIV] 25.9 µm

557

Ener

gy A

bove

Gro

und

(K)

[NII] 121.9 µm

[NII] 205.2 µm

189

70

[CI] 370.4 µm

[CI] 609.1 µm

63

39

4p: O0

3P0

3P2

3P1

327

228 [OI] 63.18 µm

[OI] 145.5 µm

[NIII] 57.4 µm

251 91

[CII] 157.7 µm

FP = 11.3 eV FP = 0, IP = 13.62

Utility: Ionized Gas Regions  Density tracers

  Far-IR fine-structure lines have Einstein A coefficients ∝ ν3  Collision rate coefficients are typically ~ qul ~ 3×10-8 cm3/sec for

e- impact excitation (and qul ~ 2×10-10 H/H2 impact excitation) ∴ since ncrit ~ A/qul we have ncrit ∝ ν3

  Furthermore the emitting levels lie far below gas temperatures ⇒ line ratios T-insensitive probes of gas density

0

3P2

3P0

3P1

[OIII] 51.8 µm

[OIII] 88.3 µm

442

163

[NII] 121.9 µm

[NII] 205.2 µm

189

70 ncrit ~ 48 cm-3

ncrit ~ 310 cm-3

Utility: Ionized Gas Regions  Hardness of the ambient radiation field

 Within an HII region, the relative abundance of the ionization states of an element depend on the hardness of the local interstellar radiation field. For example:

O+++ (54.9 eV), O++ (35.1 eV), O0 (<13.6 eV)

N++(29.6 eV), N+ (14.5 eV)

AGN

O8

O7.5

B0

Neutral ISM

Line

Rat

io

Teff (1000 K)

Line ratios between ionization states determine Teff

B2V O7.5V

Neutral Gas Lines

  [CII] 158 µm: 11.3 eV to form C+ ⇒  Arises from both neutral and ionized gas clouds

 Level is only 91 K above ground ⇒  Dominant coolant of atomic ISM

 Warm neutral medium (WNM): T ~ 5,000 K, n ~ 0.5 cm-3  Cold neutral medium (CNM): T ~ 100 K, n ~ 30 cm-3

  Important coolant for diffuse ionized clouds  Warm ionized medium (WIM): T ~ 8,000 K, n ~ 0.5 cm-3

 ncrit (neutrals) ~ 2800 cm-3 ⇒ also might expect [CII] to be an important coolant for photodissociated surfaces of molecular clouds:

n ~ 300 cm-3 5

[CII] and [OI]  O0 has same ionization potential as H ⇒ only found in

neutral gas regions  Two level system ⇒ probes n ~ few 103 to 106 cm-3

  [OI] 63 a bit problematic due to optical depth effects  (long) known to be optically thickish in emission (Stacey

et al. 1983)  Known to be self-absorbed by lower excitation gas (e.g.

Poglitsch et al. 1996)  High ncrit ⇒ emission traces dense PDRs

  [OI] 146 µm line a better probe than the 63 µm line?  Near [CII] where line ratio traces physical conditions ⇒

can detect with same system at same z  “Only” 6 times weaker than [CII]

6

Another Problem…

  [CII] line is ubiquitous – how to separate out fractions from PDRs WNM, CNM, and WIM?  Has been shown locally

that WNM and CNM contributions are small relative to PDRs – not enough column at these modest densities.

 However, collisions with e- stronger ⇒ can get substantial fraction from WIM

7

A solution: [NII] Studies   [NII] found only in HII regions  Formation energy of N++ and

C++ similar ⇒ N+ and C+ found in similar ionization zones

  [NII] 205 µm and [CII] line have same ncrit for e- impact excitation ⇒ line intensity ratio only dependent on relative C/N abundance ratio

 Carina: observed ratio ~ 9.1, ionized ratio ~ 2.4 ⇒ 75% from PDRs

8

Carina: Oberst et al. 2006

Neutral Gas Emission 9

Photodissociation Regions

Molecular cloud collapses, forming stars.

Ionized Hydrogen (HII) regions surrounding newly formed stars.

Photodissociation regions form where far-UV (6-13.6 eV) photons impinge on neutral clouds

10

Structure of the PDR   FUV flux onto neutral clouds   Thin HII/HI interface absorbs

the Lyman continuum flux   Structure depends on G/n:

 G0 [1.6×10-3 erg cm-2 s-1] ~ 1 (local) to 105 (Orion)

 n ~ 0.5 (WNM); 30 (CNM); 102.5-106 (GMCs)

  HI layer to AV~1-2: H2 interface   C+ to AV ~ 3 determined by

dust extinction of FUV photons   C+/C/CO transition   O to AV ~ 10   Heating: photo-electric effect   Cooling: fine-structure lines

[CII] and [OI]

Hollenbach and Tielens, Rev. Mod. Physics 71, 173 (1999)

PDRs are Important

By Mass  GMC PDRs

 Milky Way ~ 10%  Starbursters ~ 50%

 Atomic Gas ~ 50%  Cold neutral medium

(aka Atomic Clouds)  Warm neutral medium

By Energetics   [CII] line often brightest

single line from star forming galaxies

  [OI] similar  Also important

  [SiII] 34.8 µm  H2 rotational   [CI] 370 and 609 µm  CO rotational lines

 FIR continuum also largely arises from PDRs 11

158 µm [CII] in Star Forming Galaxies   [CII] brightest line in most star

forming galaxies with a luminosity ratio R ≡ L[CII]/LFIR ~ 0.1% to 1%

 Bi-modal [CII]/CO(1-0) correlation   [CII]/CO(1-0) ~ 4100 for starburst

galaxies and Galactic starformation regions

  [CII]/CO(1-0) ~ 1200 for non-starburst galaxies and Galactic GMCs

 Emission interpreted within PDR paradigm

12

  [CII]/CO/FIR all arise from PDRs on the surfaces of molecular clouds exposed to far-UV radiation from stars

 G0 ~ 50-2000, n~103-104 cm-3: Lower G0 appropriate for non-starburst galaxies

Stacey et al. 1991

The “[CII] Deficit” Two studies with the ISO LWS found groups of galaxies with

low [CII]/FIR continuum ratios:  30 Normal starforming galaxies (Malhotra et al. 1999)

 Galaxies with warmer 60/100 um colors and increasing FIR/Blue ratios – those with higher star formation rates – sometimes had R ~ 2×10-4 – a factor of ten lower than the typical values

  Interpreted as higher FUV fields (G0/n)

 ULIRG Galaxies (Luhman et al. 1998, 2003)   15 ULIRG galaxies studied and found 80% had [CII] deficits: the

average value was R ~ 6×10-4, or 5 times lower than “starburst” value

  Interpreted as most likely due to an additional source of FIR radiation, most likely dust bounded ionization regions – those HII regions with high ionization parameters, U ≡ Q/(4πr2nec) = NHIIαB/c

13

The [CII] and [OI] Line Trace the FUV Radiation Field Strength

 ~0.1 and 1% of the incident far-UV starlight heats the gas, which cools through far-IR line emission of [CII] and [OI] 63 µm

 The efficiency of gas heating is a function of n and G  As G rises at constant n, grain charge builds up,

lowering the excess KE of the next photo-electron  This is mitigated by raising n, enabling more

recombinations, so that the efficiency ~ G0/n  Most of the far-UV comes out as FIR continuum

down-converted by the dust in the PDRs  Therefore, the ([CII]+[OI]) to FIR ratio measures the

efficiency, hence G0/n. The combination yields both G and n, since the [CII]/[OI] ratio is density sensitive.

14

The [CII] and [OI] Line Trace the FUV Radiation Field Strength

 At a given density, the ratio R ≡ [CII]/FIR uniquely yields G

  This single-line utility holds until G/n gets small and column of [CII] determined by ion-eq.

 R ∝ 1/G since [CII] line saturates:  Above G/n ~ 3×10-3 C+ column

determined by dust extinction ⇒ ~ constant

  If n > ncrit weak n dependence   If G > 30 then Tgas > 91 K ⇒ weak

G dependence ⇒ Nearly constant I[CII], but FIR

grows with G ⇒ R ∝ 1/G  R is smaller for larger G (e.g. Orion

HII region/GMC interface) 15

[CII]/far-IR continuum luminosity ratio vs. density for various G (from Kaufman et al. 2003).

High z [CII]   IRAM 30 m telescope used

for first detection at high z J1148+5251 QSO @ z=6.42  Low R (2 × 10-4) ~ local

ULIRGS  PDR Model: G ~ 103.8, n =

105 cm-3

 Star-formation, 1500 - 3000 per year

 Subsequently resolved (PdB):  750 pc radius  SFR ~ 1000 M/yr/kpc2

(Walter et al. 2008) 16

Maiolino et al. 2005

Walter et al. 2008

High z [CII]

17

Reported in three other systems from the ground

 BRI 1202-0725 QSO @ z = 4.7 (Iono et al 2004) – R ~ 4 × 10-4 high G/n

 BRI 0912-0115 QSO @ z = 4.43 (Maiolino et al. 2009) R ~ 1.3 × 10-3  µ ~ 2.5 to 8 ⇒ LFIR ~ 4 × 1012 L  “Bright line suggests a different

star formation mechanism at high z compared with local high luminosity galaxies”

 BRI 1335-0417 QSO @ z = 4.41 (Wagg et al. 2010) R ~ 5.3 × 10-4

Iono et al. 2006

Maiolino et al. 2009

Wagg et al. 2010

18

The Redshift (z) and Early Universe

Spectrometer: ZEUS

S. Hailey-Dunsheath Cornell PhD 2009

 Submm (650 and 850 GHz) grating spectrometer ◊R ≡ λ/Δλ ~ 1200 ◊ BW ~ 20 GHz ◊ Trec(SSB) < 40 K

⇒ Limiting flux (5σ in 4 hours) ~ 0.8 to 1.1 ×10-18 W m-2 (CSO) ⇒ Factor of two better on APEX

 Data here from ZEUS – single beam on the sky  Upgrade to ZEUS-2 a ◊ 6 color (200, 230, 350, 450, 610, 890 µm

bands); ◊ 40 GHz Bandwidth ◊ 10, 9, & 5 beam system

Design Criteria   Choose R ≡ λ⁄Δλ ~ 1000

optimized for detection of extragalactic lines

  Near diffraction limit:   Maximizes sensitivity

to point sources   Minimizes grating size

for a given R   Long slit in ZEUS-2

  Spatial multiplexing   Correlated noise

removal for point sources

  Choose to operate in n = 2, 3, 4, 5, 8, 9 orders which covers the 890, 610, 450, 350, 230 and 200 µm windows respectively

ZEUS Windows

2 3 4 5

ZEUS spectral coverage superposed on Mauna Kea windows on an excellent night

500 400 600 700 800 300 Wavelength (µm)

ZEUS Redshift Coverage  ZEUS detects [CII] from z ~

1.1 to 2.1 (350 and 450 µm windows) 40% of SMGs

Covers the peak in the star-formation rate per co-

moving volume  ZEUS-2 expands [CII]

coverage to z ~ 0.25 to 5: Straddles the peak tracing the

star formation history of the Universe from 11 Gyr ago to the current epoch

 Other lines accessible as well:

ZEUS [CII] Windows

Blain et al. 2002, Phys. Rep., 369, 111

[OI] 63 µm: PDRs, n, UV field strength

[OIII] 88 µm: UV field hardness

[NII] 122 & 205 µm: ne tracer, [CII] from HII regions

ZEUS-2 ZEUS-1

First Intermediate z Detection: MIPS J1428 @ z = 1.32 w/ ZEUS/CSO

 Moderately lensed (µ→ 10) star forming galaxy Spitzer Bootes field (Borys et al. 2006): LFIR~1.3×1013 L

 Detected March 2008  Surprisingly bright: L[CII] ~ 5.4 ×1010 L   L[CII] /LFIR ~ 4.2×10-3  Modeled together with CO lines and

FIR:  G ~ 500-1000  n ~ 5000-104 cm-3

  ISM parameters similar to M82  Extended (kpc-scale) starburst: SFR

~ 2000 M/yr 21

  Hailey-Dunsheath, S. 2009 PhD (Cornell)   Hailey-Dunsheath et al. ApJ 714, L163 (2010)

([CII]+[OI]) / FIR

[CII]

/ [O

I] Herschel/ PACS Detection of [OI] in

MIPS J1428  Our [CII] detection gives

a good constraint on G0, but...

  [CII] and [OI] jointly cool PDRs  Sum is much better

estimate on cooling   [CII]/[OI] ratio is density

sensitive  Constrains G0 and n

much better.  Somewhat lower

excitation  G0 → 200  n →1000

Sturm et al. 2010

First Herschel [CII] Results: SMM J2135

@ z = 2.3  Highly lensed (µ ~ 32) submm

galaxy  Also bright line:

  [CII]/FIR ~ 2×10-3  Similar modeling as for MIPS

J1428 with CO lines:  G ~ 1000  n ~ 5000

  Indicating again, a very extended starburst

 Possible detection of [OI] line constraining PDR parameters (Danielson et al. 2010)

23 Ivison et al. A&A Herschel (2010)

A 158 µm [CII] Line Survey of the Galaxies at z ~ 1 to 2: An Indicator

of Star Formation in the Early Universe

G.J. Stacey, S. Hailey Dunsheath, C. Ferkinhoff, T. Nikola, S.C. Parshley, D.J. Benford, J.G. Staguhn,

& N. Fiolet

Stacey et al. ApJ 724, 957

24

ZEUS/CSO z = 1 to 2 [CII] Survey Survey investigates star formation near its peak in the

history of the Universe  First survey -- a bit heterogeneous

 Attempt made to survey both star formation dominated (SF-D) and AGN dominated (AGN-D) systems

 Motivated by detection – at the time of submission, only 4 high z sources reported elsewhere…

 LFIR (42.5<λ< 122.5 µm): 4×1012 to 2.5 × 1014 L  Report 12 new detections & 1 strong upper limit

 Split survey into SF-D, AGN-D and mixed (or poorly characterized) based on the characterization of TIR (3 to 1000 µm) luminosity in the literature– typically dominated by MIR for AGN systems

 12 new detections, 1 strong upper limit ◊ 6 SF-D ◊ 5 AGN-D ◊ 3 mixed 25

A Few Optical Images…

Flux

Den

sity

(10-

18 W

/m2 /b

in)

v (km/sec)

Source z LFIR F[CII] (L) (10-18 W-m-2)

RXJ094144 1.819 2.7E13 3.8±0.52

SWIRE J104738+59 1.954 7.5E12 2.8±0.36

J104705+59 1.958 5.7E12 1.7±0.45

SMM J1236 1.222 4.0E12 6.5±0.90

3C 368 1.130 5.1E12 5.1±0.80

MIPS 14282 1.325 1.3E13 19.8±3.0

Star Formation-Dominated

Sample

Flux

Den

sity

(10-

18 W

/m2 /b

in)

v (km/sec)

AGN-Dominated Sample

Source z LFIR F[CII] (L) (10-18 W-m-2)

PKS 0215 1.720 1.1E14 9.4±0.82

3C065 1.176 6.7E12 <0.8

PG1206 1.159 2.6E13 4.2±0.60

PG1241 1.273 5.9E12 15.3±2.2

3C446 1.403 2.5E14 42.8±6.5

Flux

Den

sity

(10-

18 W

/m2 /b

in)

v (km/sec)

Source z LFIR F[CII] (L) (10-18 W-m-2)

SDSS100038 1.720 5.7E12 1.74±0.40

IRAS F10026 1.176 2.0E13 14.4 ±1.1

SMM J22471 1.159 1.0E13 9.3 ±2.1

Mixed or Poorly Characterized Sample

Results: The [CII] to FIR Ratio

30

SB-D:

R = 2.9±0.5 ×10-3

AGN-D:

R = 3.8±0.7×10-4

SB-D to AGN-D ratio is ~ 8:1

Mixed – in between

[CII] Line promises to be an excellent signal for star formation at high z

L[C

II]/L

FIR

LCO/LFIR

G0=102 G0=104

n=106

G0=103

n=105

n=104

n=103 L[CII] / LCO(1-0) = 4100 n=102

Results: [CII], CO and the FIR ⇒ PDR Emission   [CII]/CO(1-0)

and FIR ratios similar to those of nearby starburst galaxies

 ⇒ emission regions in our SB-D sample have similar FUV and densities as nearby starbursters  G ~ 400-5000  n ~ 103-104

PDR Modeling

 Two sources multiple CO Lines available

 PDR parameters well constrained  SDSS J100038

 G ~ 800  n ~ 4×104 cm

 MIPS J1428  G ~ 400  n ~ 8×103 cm-3

32

PDR Modeling   Five w/ one CO line available  PDR parameters well

constrained  SMM J123634

 G ~ 500   n ~ 3 ×104 cm

 SWIRE J104738  G ~ 400   n ~ 3 ×104 cm-3

 SWIRE J104705  G ~ 600   n ~ 6 ×104 cm

  IRAS F10026  G ~ 2300   n ~ 1.6 ×103 cm-3

  3C 368  G ~ 1000   n < 5 ×104 cm-3

33

G0 from [CII] and FIR

 Seven sources have no CO lines available

 Can still confidently find Go, since we have learned from above that n ~ 103 -105 cm-3:  3C 065: G < 23,000  PG 1206: G ~10,000  PKS 0215: G ~ 7,000  3C 446: G ~ 5,000  RX J09414: G ~ 3,000  SMM J2247: G ~ 3,000  PG 1241: G ~ 150

34

How Extended are the Starbursts?  PDR models constrain G0 and n – if only [CII]/FIR we

have just G0  Since within PDRs, most of the FUV ends up heating

the dust, within PDR models, G0 ~ IFIR

 Therefore, a simple ratio IFIR/G0 yields φbeam – which then yields the physical size of the source

Inferred sizes are large – several kpc-scales  Galaxies are complex ⇒ plane parallel models are

only a first cut  Better to model distributed FUV fields (Wolfire 1990):

G0 ~ λLFIR /D3 λ << D G0 ~ LFIR /D2 λ >> D

 λ is the mean free path of a FUV photon  D = size of the emitting region 35

Extended Starbursts in the z = 1 to 2 Survey

 We use this technique to estimate source sizes based on scaling from M82:  Star formation dominated systems:

  D ~ 2 kpc for λ << D   D ~ 6 kpc for λ >> D

 Walter et al. (2009) resolved the [CII] emission from SDSS J1148 +5251 – scaling from this source for high G0 sources (e.g. PG1206) we arrive at much the same answer

 For cases where other tracers of star formation are available, our derived source sizes are typically somewhat smaller

 Star formation is extended on kpc scales with physical conditions very similar to M82 – with 100 to 1000 times the star formation rate!

36

Star Formation Evolves  Our derived physical conditions and source size contrast

markedly with local ULIRGS where:  Star formation is relatively confined (< few hundred pc)  Star formation has relatively high FUV fields – 10 times larger

than those derived here.  Growing evidence for extended starbursts at high z

 SMGs radio emission extends over kpc scales (Chapman et al. 2004), with median size ~ 5 kpc (Biggs & Ivison 2008)

 CO interferometry of SMGs shows emission over ~ 4 kpc scales (Tacconi et al. 2006, 2008)

 CO(3-2) survey of z > 1 sources suggests size scales between 3 and 16 kpc (Iono et al. 2009)

 PdBI CO interferometry of MIPS J142824 and SMM 123634: 6.3 × 4.2 & 4.3 kpc scales (Hailey-Dunsheath et al., Engle et al. in prep.)

  Likely the effect of much more massive molecular disks in the early Universe.

37

A Curious Situation

 R is significantly smaller for AGN-D systems ⇒ FUV fields significantly larger  For the same source luminosity, one would therefore

derive a smaller beam filling factor, hence smaller size  But, the AGNs are significantly more luminous ⇒

derive similar source size:   D ~ 1-3 kpc for λ << D   D ~ 3-12 kpc for λ >> D  Exception is 3C065 with D < 0.6-0.9 kpc

 The AGN-D sources have much more intense, but similar sized starbursts as the SB-D sample.

38

What Causes this Dichotomy? 1. What if the FIR in AGN comes from another source?  PAH and [NeII] line studies suggest most of the FIR from

both local AGN, and z ~ 2 AGN has its origins in star formation (Schweitzer et al. 2006, Shi et al. 2007, Lutz et al. 2008)

 Relationship between the 2-10 KeV X-ray and IR luminosity from AGN: pure systems, LX-ray/LFIR~0.45 (Ruiz et al. 2007)  Using this scaling, for all but one source (PKS 0215) we

find < 5% of the FIR is associated with the AGN.  For PKS 0215 we also show the [CII] emission might

arise from an XDR (torus) associated with the engine  For other sources, we expect little [CII] from the BLR,

NLR or XDR  Conceivable that the FIR from our two blazars (PKS 0215

and 3C446) is relativistically boosted synchrotron radiation

What Causes this Dichotomy? 2. It could simply be that the z = 1 to 2 SB-D and AGN-D

systems are distinct populations tracing different evolutionary paths

40

What Causes this Dichotomy? 3. It might be there is a temporal link between the SB-D and

AGN-D sources:  Onset of starburst and AGN activity could be coeval, or  Onset of AGN activity stimulates a kpc-scale starburst   The starburst ages as AGN activity shuts down

 Young, intense starburst associated with AGN-D system evolves into less intense, but still extended SF-D systms  Consistent with deep surveys that appear to show the peak of

AGN activity slightly before the peak of star formation activity  Observational support: [OIII] emission traces O stars. With

ZEUS/CSO we detected [OIII] 88 um from two high z (2.8 & 3.9) composites (Ferkinhoff et al. 2010), and PACS/Herschel [OIII] 52 um from MIPS 1428 and the composite FSC 10214 (Sturm et al. 2010)  Ranked in order of high to low [OIII]/FIR ratio, the AGN-D

systems tend to have strong lines ⇒ with younger starburst 41

Preferred, and testable solution: [OIII] ratio is density sensitive – how extended is [OIII] region?

If AGN generated [OIII], can test with [OIV] 24 um line.

ZEUS/CSO [OIII] at High z

 O++ takes 35 eV to form, so that [OIII] traces early type stars

 Transmitted through telluric windows at epochs of interests:  88 µm line at z ~ 3 and 4 (6) for ZEUS (ZEUS-2)  52 µm line at z ~5.7 and 7.7! --- much more challenging

 Detectable in reasonable times for bright sources

42

ZEUS/CSO Detections

 Detected in in 1.3 hours of integration time on CSO – differences in sensitivity reflect telluric transmission

 Two composite systems  APM 08279 extremely lensed (µ → 4 to 90)  SMM J02399 moderately lensed (µ ~ 2.38)

43

Ferkinhoff et al. 2010 ApJ 714, L147

Characterizing the Starbust/AGN   [OIII]/FIR

 APM 08279 ~ 5.3 × 10-4; SMM J02399 ~ 3.6 × 10-3  Straddles the range (2×10-3 ) found in local galaxies

(Malhotra et al. 2001, Negishi et al. 2001, Brauher et al. 2008)

 Origins of [OIII]: APM 08279  Very few tracers of star formation available: e.g. Hα,

Paα, Paβ lines are broad and variable with luminosities 100 times larger relative to [OIII] 88 µm than for M82

 Spitzer PAH upper limit 10 × F[OIII], and expect ~ unity

Build both starburst and AGN model

AGN Origin for APM 08279?  AGN: NRL ne ~ 100 – 104 cm-3 <ne> ~ 2000 cm-3

(Peterson 1997)  For this ne range, [OIII] 5007 Å/ [OIII] 88µm ~ 1 to

100  Using this relationship and the fact that: LO3, 5007 ~

Lbol/3500 (Kaffmann & Heckman 2005)  ⇒ expect L[OIII] 88 µm ~ 1 to 100 × 1010 /µ L (fcn ne)  ⇒ all 1011 /µ L [OIII] may arise from NLR if ne > 2000

cm-3

Starburst Origin for APM 08279   [OIII]/[NII] line ratios

insensitive to ne, but very sensitive to Teff   [OIII]/[NII] 122 especially

so…  Ratio in APM 08279 > 17

based on non-detection of 205 µm (Krips et al. 2007) ⇒ Teff> 36,000 K ⇔ O9

stars  If starburst headed by

O7.5, then 35% of FIR from starburst: SFR ~ 12,000/µ M/year 46

From Rubin, R. 1985

Starburst Origin for SMM J02399

  Ivison et al. 2010 argue for starburst origin of FIR  6.2 µm PAH flux ~ [OIII] 88 µm line flux as for

starbursters  Using scaling laws and the observed Lα, [OIII] 5007

Å and 88 µm fluxes we estimate:   [OIII]/Hα (sensitive to Teff) and   [OIII] 5007 Å/ [OIII]88 µm (sensitive to ne and Te)  Yields: Teff ~ 40,000 K (O7.5) an ne ~ 300 cm-3

 SMM J02399 fit by a starburst headed by O7.5 stars, forming stars at a rate ~ 5000/µ per year.

47

PACS [OIII] 51.8 µm Detections

 PACS/Hershel detections of [OIII] 51.8 µm line from two sources

 PACS takes advantage of happy coincidence: 25.9 x 2 = 51.8

  [OIV] 25.9 µm (2nd order), [OIII] 51.8 µm (1st order)

 Line ratio very sensitive tracer of AGN

48 Sturm et al. 2010

 To form O++ takes 35.1 eV; to form O+++ takes 54 eV  Strong [OIV] requires an AGN on galactic scales – the

[OIII] 88 / [OIV] 26 ratio is as good as NeII/NeV 49

[OIV] 26 µm: AGN AGN

Starbursts

PACS/ZEUS Synergy

 Bright fine-structure lines of roughly equal luminosity for most galaxies

 ZEUS has roughly same sensitivity (W/m2) at 350 µm as PACS does from 100 to 200 µm

 Therefore, ZEUS can detect [CII] line in z = 1 to 2 interval as PACS detects the [OI] (63 µm), and [OIII] (52 and 88 µm) lines in the same redshift range.

 Even get the [OIV] 25.9 um AGN tracer (same L as [OIII]!) for free.

  Powerful diagnostics of stellar types and the presence of AGN

50

51

ZEUS [CII] Work in Perspective  Observations of [CII] at z ~ 1

to 2 characterize star formation in galaxies at the historic peak of star formation in the Universe

  ZEUS on the CSO provides a unique opportunity to explore this epoch through the [CII] line

 Approximately 40% of the submm galaxy population has redshifts such that the [CII] line falls in the 350 (z ~ 1) or 450 (z~2) µm windows

ZEUS [CII] Windows

Blain et al. 2002, Phys. Rep., 369, 111

With ZEUS-2 on CSO and APEX we can extend these studies from z ~ 4 to 0.25 -- tracing the history of star formation from its peak 11 Gyr ago to the present epoch

ZEUS-2 ZEUS-1

The Near Future: ZEUS-2   Upgrading to (3) NIST 2-d

TES bolometer arrays   Backshort tuned   5 lines in 4 bands

simultaneously   215 µm (1.5 THz)   350 µm (850 GHz)   450 µm (650 GHz)   625 µm (475 GHz)

  Imaging capability (9-10 beams)

  Simultaneous detection of [CII] and [NII] in z ~ 1-2 range

  First light in January 2011 on CSO with 400 um array only

  APEX later in 2011

spectral

CO(7-6)

[CI] 370 µm 13CO(6-5) µm

[CI] 609 µm

[NII] 205 µm

9 × 40 400 µm array

10 × 24 215 µm array

5 × 12 625 µm array

spatial

Imaging Capabilities

M51 - CO(1-0): BIMA Song (Helfer et al. 2003)

 Astrophysics   [CI] line ratio: Strong

constraints on T  13CO(6-5) line: Strong

constraints on CO opacity   [NII] line: Cooling of

ionized gas, and fraction of [CII] from ionized media

 Mapping Advantages  Spatial registration

“perfect”  Corrections for telluric

transmission coupled  Expected SNR for the five

lines comparable

12CO(7-6) 13CO(6-5)

[CI] 3P2 - 3P1

[NII] 3P1 - 3P0

[CI] 3P1 - 3P0

The Near Future: ZEUS-2

54

FS Lines and ALMA When complete, the ALMA array will be ~ 20 to 30

times more sensitive than ZEUS on APEX

55

ZEUS-2 Line Sensitivities on APEX

4E11

4E10

4E9

8E8

Band 10

4E12

4E11

4E10

8E9

Conclusions   [CII] line emission at high z

 Reveals star forming galaxies  High star formation rates with local starburst-like

physical conditions  ⇒ Extended starbursts  Extended starbursts in AGN

  [CII] from XDR?   [OIII] emission at high z

 Traces current day stellar mass function  Also can traces physical conditions of NLR – may

have detected it from APM 08279  Future with ALMA exciting – resolve sources that are

10’s of times fainter 56