Massive molecular gas flows and AGN feedback in galaxy clusters

CO(3-2)

Helen Russell (Cambridge)Brian McNamara (Waterloo), Andy Fabian (Cambridge), Paul Nulsen (CfA),

Michael McDonald (MIT), Alastair Edge (Durham), Francoise Combes (Paris), Philippe Salomé (Paris), Jeremy Sanders (MPE) and the SPT collaboration

● Introduction

– Radiative cooling in galaxy clusters

– AGN feedback

● Results

– ALMA observations of molecular gas in central cluster galaxies

– Massive extended filaments, smooth velocity gradients

– AGN-driven gas outflows

– Direct uplift of molecular clouds or cooling in situ?

● Conclusions

Outline

X-ray surface brightness peaks in cluster cores

Lewis et al. 2003, Allen et al. 2004

● 100 – 1000 M per year gas cooling?

X-ray radiative cooling time

Voigt & Fabian 2004

Optical

Coo

ling

time

(Gyr

)

Chandra X-ray

100 kpc 100 kpc

Radius (kpc)

● Searches for vast reservoir of molecular gas find less than 10% of that expected (Edge '01, Salomé + Combes '03) → residual cooling

● AGN heating replaces radiative losses → feedback loop

● Truncates galaxy growth, keeps ellipticals ‘red and dead’, M-σ relation

MS0735, McNamara et al. 2005

200 kpc

Optical, X-ray, Radio

Radio jets heat cluster gas

LX(<r

cool) (1042 ergs-1)

Pca

v (10

42 e

rgs-1

)

Rafferty et al. 2006; Birzan et al. 2004; Fabian 2012

He

atin

g

Cooling

Chandra: energy output ALMA: fuelling?

● Origin of molecular gas in BCGs?

● Is molecular gas fuelling feedback?

● Does radio-jet feedback operate on molecular clouds?

● BCGs in cool core clusters are rich in molecular gas (Edge 2001, Salomé & Combes 2003)

What is the role of molecular gas in feedback?

0.3”

0.5 kpc

● Massive filaments each ~ a few x 109 – 1010 M and 3 – 15 kpc long

Russell et al. 2016, submitted

PKS0745

Extended filaments of molecular gas

Phoenix

CO(3-2)

● CO(2-1) absorption features with ~ 5 km/s linewidth typical of GMC and infalling velocity 250-350 km/s

David et al. 2014

Filaments consist of many GMCs

NGC5044

Flu

x (m

Jy)

Velocity (km/s)Tremblay et al. 2016

A2597

Low velocities and low dispersions

● Modest velocities ±100 km/s, narrow FWHM ~100 km/s

● Gas not settled in gravitational potential

Beam size: 0.3 arcsec, 0.5 kpc Russell et al. 2016

Velocity map Velocity dispersion

PKS0745:

PKS0745

Smooth velocity gradients along filaments

● Low velocities at small radii

● Highest velocities at largest radii

● Outflow?

A1795:

Highest velocities at largest radii

McNamara et al. 2014

A1835

● 1010 M molecular flow at 200-400 km/s extending to 10kpc

A1835:

Phoenix cluster: ordered gas flow to centre

● Smooth velocity gradients and low FWHM in filaments

● Velocity gradient across nucleus with much higher FWHM

● Velocities too low for free fall in gravitational potential

CO(3-2)

Russell et al. submitted

Molecular gas not settled in the gravitational potential

● Massive filaments, low velocities → merger origin is unlikely

● Low velocities compared to stars → filaments not supported by rotation

● Highest velocities at large radii → outflow?

Fabian et al. 2003; Conselice et al. 2001; Lim et al. 2008; Salome et al. 2011

CO detections on Hα image

Perseus: Hα

Molecular gas filaments extend toward radio bubbles

0.3 arcsec

0.5 kpc

Sanders et al. 2014

Russell et al. 2016

● PKS0745: massive filaments drawn up underneath X-ray cavities and radio lobes

A1835 CO(1-0)

2kpc

● Gas filaments drawn up around radio bubble

● Interaction with cold gas in radio-mode feedback

A1835: gas flow drawn up around the X-ray cavities

6.7 kpc

1.7 arcsec

X-ray cavities

McNamara et al. 2014

Phoenix cluster: filaments encase cavities

● Phoenix hosts a ~600 M /yr starburst and a 'transition' AGN bright in both X-ray/optical and radio

● 3 x 1010M of molecular gas total with half in filaments around radio bubbles

30 kpc

CO(3-2)

Russell et al. submitted

HST

McDonald + SPT collaboration 2012

cavitypositions

● Molecular gas structure shaped by radio bubble expansion

● Direct uplift of molecular gas clouds?

– Pcav ~ 1043-45 erg/s

– High coupling efficiency required

● Rapid cooling of uplifted thermally unstable low entropy gas?

– Molecular gas coincident with soft X-ray

– Dust lanes

Direct uplift of molecular gas clouds or cooling in situ?

Phoenix

PKS0745

Phoenix

● Molecular gas structure shaped by radio bubble expansion

– Massive 109-1010M filaments drawn up around and beneath radio bubbles

Conclusions

● Molecular emission lines are narrow– Extended filaments, ordered velocity

structure– Gas not settled in gravitational potential– Circulation flow

● Radio bubbles supply large-scale heating to stabilise cluster atmospheres and lift gas in their wakes

Phoenix

PKS0745