The Evolutionary Connection Between SMGs & AGN as Probed by - - PowerPoint PPT Presentation

The Evolutionary Connection Between SMGs & AGN as Probed by Molecular Gas Excitation

Chelsea E. Sharon

Herschel Postdoctoral Fellow McMaster University Collaborators: Dominik Riechers, Jackie Hodge, Chris Carilli, Fabian Walter, Axel Weiss, Kirsten Knudsen, & Jeff Wagg

Sharon et al. 2016, ApJ, 827, 18

Reminder: Theoretical Motivation for AGN Feedback

Weller et al. (2005) Bell et al. (2003) Read & Trentham (2005): Elliptical galaxies All galaxies

–5 –10 –15 –20 6 8 10 12 14

–1) log10 (M/M )

D a r k m a t t e r

SN AGNs

Dark matter * baryon fraction

log10 N (Mpc–3 M Kormendy & Ho 2013 ARA&A

Reminder: 2 AGN Feedback Modes

• Kinetic/jet/radio/maintenance

mode

• Definitely important in

most massive galaxies

• Effects on hot halo gas

are observed

• Radiative/wind mode
• Outflows are ubiquitous
• Mostly observed in

ionized and atomic phases

• Difficult to distinguish

between AGN-driven and starburst-driven winds

Perseus Cluster (Chandra/X-ray) Rubin et al. 2013

AGN Feedback & Molecular Gas

• Recent work clearly

identifies molecular gas

• utflows in nearby AGN-

host galaxies

• Some low-z AGN have

ultra-high excitation molecular lines

• But similar observations

are challenging at peak epoch of galaxy assembly

• 1000 -500

500 1000 0.1 0.2 0.3 Velocity [Km/s]

• 1000 -500

500 1000 0.01 0.02 0.03 Velocity [Km/s]

Feruglio et al. 2010 Cicone et al. 2014 (see also Fiore et al. 2017)

If AGN quench star formation, we should see signs in the molecular gas

AGN Feedback & Molecular Gas

• Recent work clearly

identifies molecular gas

• utflows in nearby AGN-

host galaxies

• Some low-z AGN have

ultra-high excitation molecular lines

• But similar observations

are challenging at peak epoch of galaxy assembly

• 1000 -500

500 1000 0.1 0.2 0.3 Velocity [Km/s]

• 1000 -500

500 1000 0.01 0.02 0.03 Velocity [Km/s]

Feruglio et al. 2010

5 10 15 20

Jup

10 20 30 40 line flux [10-17 W m-2]

interferometric

JCMT ext

FTS map FTS deep PACS

Spinoglio et al. 2012 (see also Rosenberg et al. 2015)

If AGN quench star formation, we should see signs in the molecular gas

A Promising Connection? Low-J CO Excitation

• Based on radiative transfer modeling, r3,1 is a good

indicator of the molecular gas physical conditions # Phases r3,1 At high-z: Single phase, thermalized r3,1=1 AGN host galaxies r3,1=1

(Riechers et al. 2011)

Multi-phase r3,1<1 Submillimeter galaxies (SMGs) r3,1=0.6

(e.g., Swinbank et al. 2010; Harris et al. 2010; Ivison et al. 2011; Riechers et al. 2011)

8 5

r

New(-ish) Survey: Total Molecular Gas Census in z~2 SMGs & AGN

• VLA CO(1–0) observations of all

z~2 SMGs & AGN with existing CO(3–2) detections

• Goal: 5–7σ detections for each

(although not the case in practice)

• 13 new detections + 1 limit
• +15 literature sources

➡ 15 SMGs (10 lensed/5 unlensed) ➡ 14 AGN (9 lensed/5 unlensed)

5 spectrally resolved 3 >5σ detections 5 3–5σ detections 1 upper limit

27.3s 27.4s 27.5s 27.6s 27.7s 27.8s

RA (J2000) 09

50″ 52″ 54″ 56″ 58″

DEC (J2000) +05° 50′

RXJ0911+0551 (QL) z=2.796 σ=0.008 Jy km s

−1 beam −1

Δvint=200 km s

−1

CO(1−0)

17.0s 17.2s 17.4s 17.6s 17.8s 18.0s

RA (J2000) 10

0″ 5″ 10″ 15″

DEC (J2000) +43° 45′

HS1002+4400 (Q) z=2.101 σ=0.055 Jy km s

−1 beam −1

Δvint=1200 km s

−1

CO(1−0)

34.8s 35.0s 35.2s 35.4s 35.6s 35.8s

RA (J2000) 22

30″ 35″ 40″ 45″

DEC (J2000) +00° 15′

J22174+0015 (S) z=3.099 σ=0.020 Jy km s

−1 beam −1

Δvint=1200 km s

−1

CO(1−0)

37.4s 37.6s 37.8s 38.0s 38.2s 38.4s

RA (J2000) 04

35″ 40″ 45″ 50″

DEC (J2000) +05° 34′

MG0414+0534 (QL) z=2.639 σ=0.020 Jy km s

−1 beam −1

vint=580 km s

−1

CO(1−0)

Is there a difference?

Old Distribution New Distribution

AGN SMGs 0.0 0.5 1.0 1.5 2.0 2.5 r3,1 2 4 6 8 N 0.0 0.5 1.0 1.5 2.0 2.5 r3,1 2 4 6 8 N AGN SMGs 0.0 0.5 1.0 1.5 2.0 2.5 r3,1 0.2 0.4 0.6 0.8 1.0 Cumulative Fraction AGN SMGs 0.0 0.0 0.5 1.0 1.5 2.0 2.5 r3,1 0.2 0.4 0.6 0.8 1.0 Cumulative Fraction AGN SMGs 0.0

No!

What about the line FWHMs?

200 400 600 800 1000 FWHM (km s–1) 1 2 3 4 5 6 N CO(1−0) CO(3−2) 0.6 0.8 1.0 1.2 1.4 FWHM3–2/FWHM1–0 1 2 3 4 5 N SMGs AGN 200 400 600 800 1000 FWHM (km s–1) 0.0 0.2 0.4 0.6 0.8 1.0 Cumulative Fraction CO(1−0) CO(3−2) 0.6 0.8 1.0 1.2 1.4 FWHM3–2/FWHM1–0 0.0 0.2 0.4 0.6 0.8 1.0 Cumulative Fraction SMGs AGN

No!

FWHMs FWHM Ratios

Why did the difference disappear?

• It’s easy to excite CO(3–2) in dense star forming regions
• Issues with AGN/star formation content:
• AGN are largely literature sources (as per initial r3,1

detections) and not selected in a physically motivated way

• These AGN have star formation (IR bright), and relative

strength of the AGN (compared to SF) is not quantified

• Original samples were well-characterized sources, clear on

AGN content, and expanded samples are less clear

• Some galaxies may be mis-classified, or their central AGN

is deeply buried in dust

• Differential lensing may affect the observed excitation

conditions

What else can we do?

N=1 for high-z and/or U/LIRGs N=1.2 for low-z IR-bright (Y03)

• Correlation hunting between r3,1 and other

parameters doesn’t turn up anything interesting

• Do see correlation between r3,1 and star formation

efficiency, as previously found at low z

• No significant excitation dependence in integrated

Schmidt-kennicutt relation, as found at low z

100 200 300 400 500 LFIR/LʹCO (L (K km s–1 pc2)–1)

AGN, unlensed AGN, lensed SMG, unlensed SMG, lensed local IR-bright galaxies (Y03) local U/LIRGs (P12+G14)

0.0 0.5 1.0 1.5 2.0 2.5 r3,1

AGN, unlensed AGN, lensed SMG, unlensed SMG, lensed local IR-bright galaxies (Y03)

• ther

CO(1–0) CO(3–2) local U/LIRGs (P12+G14)

log(LʹCO/(K km s–1 pc2)) log(LFIR/L ) 7 8 9 10 11 12 9 10 11 12 13 14

Cosmic Eye cB58 FLS 3 (z=6.34)

What’s next?

• Argument still holds: If AGN

feedback is directly affecting star-forming gas, we should see signs of that in the gas physical conditions.

• Need a way to disentangle

the relative strength of AGN’ and star formation’s effects on the ISM. →Survey of gas excitation in unlensed systems where we have good (mid-) IR diagnostics

• f the AGN strength.

Kirkpatrick et al. 2015

New Survey: Multi-J CO in High-z Galaxies with IRS spectra

• Started survey at the VLA and got CO(1–0) for 11 of 12

sources so far (6 completed, above)

• Note: these only had IRS redshifts prior to CO(1–0)
• NOEMA observations of higher-J lines in a sub-sample

have just completed…

Summary/Concluding Thoughts

• No systematic difference in CO(3–2)/CO(1–0) line

ratios in SMGs and IR-bright AGN host galaxies

• No difference in CO(1–0) and CO(3–2) line FWHMs
• No new trends in CO(3–2)/CO(1–0) with other galaxy

properties (reproduces low-z correlation with SFE)

• Lack of difference largely because high SFRs can easily

excite high CO(3–2)/CO(1–0) ratios and IR-bright AGN also have high SFRs

• Samples w/ better handle on AGN strength at long

wavelengths will give us a better sense of AGN-SF connection and are in progress Open Question: What role, if any, do AGN play in the evolution of dusty starbursts like SMGs?