Dense gas really matters.

Transcription

1 Dense gas tracers and star formation laws: Multiple transition CS survey in nearby active star-forming galaxies Zhi-Yu Zhang U. Edinburgh/ESO! Dense gas really matters. Collaborators: 19 Dec. 2014, Guangzhou, Zhi-Yu Zhang Yu Gao (PMO) Padelis Papadopoulos (Cardiff) Thomas Greve (UCL) Christian Henkel (MPIfR) Junzhi Wang (SHAO) Yinghe Zhao (PMO/IPAC) Manolis Xilouris (NOA) Rob Ivison (ROE/ESO) Karl Menten (MPIfR) et al.

2 Outline Background Gas tracers and Star formation Star formation laws! Surveys and Results Multiple-J CS surveys in galaxies Star formation vs. dense gas emission!! MALATANG and more. Summary

3 Which gases are forming stars? IC 342 HI (atomic gas) THINGS 1 kpc

4 Which gases are forming stars? IC 342 HI (atomic gas) 12 CO J=1-0 (molecular gas) THINGS NRAO 12m 1 kpc

5 Which gases are forming stars? IC 342 HI (atomic gas) 12 CO J=1-0 (molecular gas) THINGS NRAO 12m Spitzer 70um IR emission (star formation) 1 kpc On kpc scales, SFR is more related to H2 gas, rather than to HI.

6 Which gases are forming stars? IC 342 HI (atomic gas) 12 CO J=1-0 (molecular gas) THINGS NRAO 12m Spitzer 70um IR emission (star formation) GRS 13 CO Galactic Ring Survey (GRS) 13 CO J=1-0 1 kpc On kpc scales, SFR is related to H2 gas,rather than HI

7 Which gases are forming stars? - Galactic view GRS 13 CO J=1-0 Extended on ~ 10 pc scales Low volume density ~ cm pc

8 Which gases are forming stars? - Galactic view MSX 20 μm Compact on ~ sub-pc scales 10 pc

9 Which gases are forming stars? - Galactic view GRS CS J=2-1 Compact on ~ sub-pc scales High volume density ~ cm pc

10 Star Formation Laws: Gas - SFR relations SFR N=1.4 N=1.0 N=1.0 Which tracers are tracing the star forming gas? HI+H2 Mgas/tff H2(dense) Gas Mass Kennicutt 1998 Krumholz et al Gao & Solomon 2004

11 Tracers of Physical Conditions in Molecular Clouds High excitations Shocks, XDR etc. GMCs Dense Cores Genzel 1992

12 ense gas tracers When n(h2) > ncrit: ncrit > 10 4 cm -3 ncrit(hcn) : 10 4 ~10 7 cm -3 Collisional excitation dominant. Easily be thermalised. ncrit(hco + ) : 10 4 ~10 6 cm -3 ncrit(co) : 10 2 ~10 5 cm -3 ncrit(cs) : 10 4 ~10 6 cm -3 HCN : IR-pumping, XDR, chemistry on Tkin. e.g. Weiss et al. 2008; Graci-Carpio et al. 2006; Lintott & Viti 2006; Baan et al HCO + : Shock, ionisation fields, etc. e.g. Dickinson et al. 1980; Dickmann et al. 1992; Papadopolous et al High-J CO: Warm gas. CS : Weak (~1/4 of HCN intensity), chemically stable? e.g. Charnley 1997; Martín et al. 2008;2009

13 Simulations of star formation laws higher transitions/densities have lower slope indices? Slope Juneau09 Transition Aul/Cul Narayanan et al Juneau et al. 2009

14 L gas-lir correlations -- CO 1-0 (ncrit~ 4 x10 2 cm -3 ) Slope~1.4 Gao & Solomon 2004 L gas -- Mgas LIR -- SFR

15 L gas-lir correlations -- HCN 1-0 (ncrit~ 2 x10 5 cm -3 ) Slope=1 Gao & Solomon 2004 L gas -- Mgas LIR -- SFR

16 L gas-lir correlations -- HCO (ncrit~ 1 x10 6 cm -3 ) Slope=0.8 Juneau et al Juneau et al L gas -- Mgas LIR -- SFR

17 L gas-lir correlations -- HCN 3-2 (ncrit~ 5 x10 6 cm -3 ) N~0.7 Bussmann et al L gas -- Mgas LIR -- SFR

18 Galactic CS & HCN studies 1.03(0.05) 1.17(0.06) 0.88(0.06) 1.05(0.05) 0.81(0.04) 1.03(0.05) 1.05(0.05) 0.81(0.04) Wu et al. 2010

19 Issues in observational results IR pumping? Chemistry? Stable tracers need. IR size > beamsize Either to map gas emission or to match IR with beam Bussmann et al To test the above models in galaxies, multiple transition surveys of chemically clean dense tracers, e.g. CS lines, are needed.

20 Sample Selection: 1. IRAS Revised Bright Galaxies sample (IRAS RBGs, Sanders 2003). Flux cutoff: F100um >100 Jy, F60um >50 Jy.! 2. Rich detections of CO and HCN lines.! 3. A large range of LIR, and galaxy types: Nearby normal galaxies, starburst, LIRGs, and ULIRGs. ~ 50 galaxies are selected

21 Multiple-J CS survey ~ 280 hours in total Multiple transitions (J=1-0 to 7-6) of CS lines towards ~ 40 nearby normal galaxies, starburst, and (U)LIRGs CS J= 2-1/3-2/5-4 IRAM 30m CS J= 5-4 SMT(HHT) 10m CS J= 5-4/7-6 APEX 12m 80 hours 50 hours 40 hours 2009 ~ CS J= 1-0 GBT and the EVLA 60 hours ~40 hours

22 Samples and Detections CS J=1-0 : 20/24 galaxies CS J=2-1 : 41/47 galaxies CS J=3-2 : 30/41 galaxies CS J=5-4 : 21/40 galaxies + CS detections in literatures CS J=7-6 : 11/20 galaxies HCN/HCO + J=4-3 simultaneously C 34 S J=2-1: 5 detections Mrk 231 NGC 4418 NGC 1068

23 L gas-lir (small targets) CS J=1-0 ncrit~ 1x10 4 cm -3

24 L gas-lir (small targets) CS J=1-0 to J=5-4 N=0.95 N=1.02 N=1.04 N=0.96

25 Beam matching photometry for extended targets NGC 891 Beam Whole Herschel 100 μm Bussmann et al LSD = RSD LTIR(IRAS) RSD=Fbeam/Ftotal varies at different bands Assuming whole galaxy share one IR SED.

26 L cs-lir correlations Beam matching correction All linearly correlated

27 L cs-lir correlations ~ 8 orders of magnitude N=1 galaxies ncrit~ 1x10 5 cm -3 Wu et al Galactic cores

28 L cs-lir correlations ~ 8 orders of magnitude N=1 galaxies ncrit~ 2x10 6 cm -3 Wu et al Galactic cores

29 L cs-lir correlations ~ 8 orders of magnitude N=1 galaxies ncrit~ 5x10 6 cm -3 Wu et al Galactic cores Zhang et al. 2014

30 HCO + J= observed simultaneously with CS J=7-6 ncrit~ 2x10 6 cm -3 Zhang et al Zhang et al. 2014

31 HCO + J= observed simultaneously with CS J=7-6 Slightly Super-Linear! N~0.5 prediction ncrit~ 2x10 6 cm -3 Zhang et al Zhang et al. 2014

32 HCN J= observed simultaneously with CS J=7-6 ncrit~ 1x10 7 cm -3 Zhang et al Zhang et al. 2014

33 HCN J= the highest ncrit tracer N~0.5 ncrit~ 1x10 7 cm -3 Narayanan s prediction Zhang et al Zhang et al. 2014

34 Dense gas tracers with ncrit ~ cm -3 CS Slope HCO HCN 4- Juneau et al Aul/Cul Dense gas tracers have linear correlations irrespective to ncrit, universally over 8 orders of luminosity magnitudes.

35 Does time scale matter? -- For Dense gas: probably No. SFR fh2 : H2 fraction ϵff : constant, dimensionless measure of SFR Krumholz et al Mgas/tff If LIR = (L dense) N /tff, N will decrease with ncrit. This will be contradictory to our observed results.

36 Short summary 1) Dense molecular gas (n(h2)~> 10 4 cm -3 ) is the star-forming gas.! 2) How much dense gas, how much star-formation linear correlations.! 3) L dense-lir universally stays linear from Galactic cores to galaxies, irrespective to critical density, once it is > 10 4 cm -3.!

37 The other half of the story Either to map gas emission or to match IR with beam Bussmann et al. 2008

38 MALATANG co-pi: Yu Gao, Zhiyu Zhang,Thomas Greve JCMT 390 hrs large program! Mapping HCN/HCO+ J=4-3 in ~20 nearby star-forming galaxies.! Synergy with Herschel FTS high-j CO and excitation modelling.! Characterising the physical/chemical conditions and excitations of the SF units probed by HCN/HCO+

39 Dense gas emission on disks and arms.

40 Different mode of star formation on disks? galactic centres M33 M51 M31 Chen et al Usero et al MALATANG will give the answers.

41 Synergy the HCN/HCO+ SLEDs with CO Papadopoulos

42 Thank you!

43 Background: Gymnastic music ALMA will be helpful!

44 Stars are forming in dense molecular gas cores Diffuse atomic gas HI, the gas reservoir for molecular clouds, And the supply for future star formation. PDRs GMCs: n(h2) ~ cm -3 Tkin ~ K D ~ pc Dense cores: n(h2) ~ cm -3 Tkin D ~ K ~ pc self-gravity bound from Y. Gao s talk

45 Backup Slides

46 LVG+ML/Bayesian Modelling with dense gas tracers and CO Papadopoulos

47 Model high-j CO using LVG results of HCN (NGC 6240) ~60-70% of the molecular gas is in dense gas phase. The thermal state of molecular gases can not be maintained by FUV from PDRs. Detailed LVG analysis will be done for the whole sample.

48 Caveats L dense is a first order approximation of Mdense. Detections of high ncrit lines do not necessarily mean that the gas densities are above ncrit, because they can be subthermally excited.! Analysis on excitation conditions is needed.

49 Surface density correlation of HCN -10

50 HCO + J=1-0 Ma et al. 2013

51 Fitting results Intercept vs. J CS 7-6 CS1-0 sub-linear slope indices for uncorrected targets linear correlations for point targets and beam matched targets

52 Aperture Correction -- beams are small : Parameters of the photometry. CS2-1 (25 ) CS 3-2 (17 ) Source name 24Ratio 24Apercor 70Ratio 70Apercor 24Ratio 24Apercor 70Ratio 70Apercor MAFFEI IC NGC M Aperture NGC Convolution Photometry NGC NGC correction Final flux NGC NGC NGC NGC NGC NGC NGC NGC NGC NGC ARP UGC NGC MCG MRK UGC Fluxbeam=Ftotal x Rbeam/total x Apercorr UGC Mrk IRAS IRAS IRAS VIIZW IRAS K blackbody PSF

53 Star Formation Law (Units) L IR (Mʘ yr -1 pc 2 ) Σ SFR SFR (Mʘ IR (Mʘ Σ yr SFR/tdyn luminosity yr -1-1 ) pc (Mʘ -2 ) yr -1 pc -2 yr -1 ) Global SFL SFR Surface Density without size measurement The one we want. Gao & Solomon 2004 Bussmann Kennicutt Kennicutt et al Krumholz Graciá-Carpio Bigiel et et al. al. et al Wang et al Schruba et al Σ M L gas/tdyn Gas (K (Mʘ) kms (Mʘ pc -1-2 pc pc ) -2 Surface 2 ) yr -1 ) Density Gas luminosity

54 Extended CS emission on the disk MAFFEI 2 Spitzer 24 μm CS 2-1(IRAM 30m)

55 ense gas tracers Critical Density: Rotational transitions of heavy molecules HCN, HCO+, CS, high-j CO etc. Dense gas tracer: ncrit >10 4 cm -3 Dense Cores CO 1-0 CO 2-1 CO 3-2 CO 5-4 CO 4-3 Giant Molecular Clouds hot cores XDR etc. Except for abundance and excitation,! molecular emissions can be influenced by:! radiative pumping, chemistry, electron density, shock dissociation, etc.

56

57

58 HCO + deficient in extreme conditions?? Higher slopes for HCO + (only) in galaxies. Gracia - Carpio et al. 2006, 2008; Imanishi et al Linear in Galactic cores, e.g., Ma et al HCO + is an ionic molecule. HCO + + e CO + H High radiation fields in ULIRGs X-ray / Cosmic Rays => high n(e) Papadopoulos et al Shock environment Shocks produce electron-rich outer layers Xie et al. 1995

59 Why slopes matter? Different SFE Which gases are forming stars? 22 N= SFR N= N= Gas Mass

60 LIR-LCO relations ULIRGs low/high-z Greve+14

61 LIR-LCO relations ULIRGs low/high-z Greve+14

62 Theoretical works 1) Krumholz et al. (2007): ncrit < nave: slope ~ 1.5 e.g., CO 1-0 ncrit > nave: slope ~ 1 e.g., HCN 1-0 2) Narayanan et al. (2008): Sub-thermal excitation. Slope decreases with ncrit. slope nave slope ncrit N=1 N=1 3) Lada et al. (2012): Linear slope for lines with ncrit > 10 4 cm -3 SFR is only related to Mdense. K-S law slope is related to Mdense fraction. slope ncrit 10 4 cm -3 ncrit N=1

63 Low-J CO: gas not all forming stars. mid-&high-j CO: Mid-J CO: star forming gas. High-J CO: extra heating mechanism. Greve et al Lu et al. 2014

64 Slope (α) vs. J Greve+14

65 Slope (α) vs. J Dense and cold gas tracers have linear correlations irrespective to ncrit. This is valid universally over 8 orders of luminosity magnitudes. Excitation conditions next step Greve+14

66 DeMoGas HerCULES sample Full CO ladders (from J=1-0 to 13-12) 13 CO ladders Multiple molecules (HCN/HCO+/CS/etc.) Multiple transitions The most complete dataset of dense gas tracers in nearby U/LIRGs. Manolis Xilouris Ioanna Leonidaki Padelis Papadopoulos Paul van der Werf Thomas Greve Zhi-Yu Zhang Panos Boumis Alceste Bonanos