1. Atomic and Molecular Gas in Galaxies Mark Krumholz UC Santa Cruz The EVLA: Galaxies Through Cosmic Time December 18, 2008 Collaborators: Sara Ellison (U. Victoria) Chris Matzner (U. Toronto) Chris McKee (UC Berkeley) Xavier Prochaska (UC Santa Cruz) Jonathan Tan (U. Florida) Jason Tumlinson (STScI)

2. Stars Do Not Form in Gas SINGS + GALEX + THINGS + SONG (animation borrowed from N. Gnedin)

3. Stars Form in Molecular Gas The SFR in a galaxy correlates with its molecular gas surface density. Atomic gas is important primarily for producing molecules. SFR vs. surface densities of HI (blue asterisks) and H 2 (black and green triangles) in M51a (Kennicutt et al. 2007)

4. Implications  The relationship between HI content and star formation is not straightforward, particularly at low surface density  A theory of star formation must be able to predict the atomic and molecular fractions in a galaxy’s ISM from physical principles

5. Anatomy of An Atomic- Molecular Complex Molecules reside in giant molecular clouds (GMCs) that are the inner parts of atomic-molecular complexes The outer parts are dissociated by interstellar Lyman-Werner photons Goal: compute HI and H 2 mass fractions

6. Dissociation Balance in Atomic-Molecular Complexes (Krumholz, McKee, & Tumlinson 2008a) Idealized problem: spherical cloud of radius R, density n, dust opacity  d, H 2 formation rate coefficient R, immersed in radiation field with LW photon number density E 0 *, find fraction of mass in HI and H 2. Decrease in radiation intensity = Absorptions by H2 molecules + dust grains The basic equations for this system are chemical equilibrium and radiative transfer. Formation on grains = Photodissociation

7. Calculating Molecular Fractions To good approximation, solution only depends on two dimensionless numbers: Approximate solution: Top: analytic solution for location of HI / H 2 transition vs. exact numerical result Bottom: H 2 volume fraction vs. ,  R

8. Shielding Layers in Galaxies (Krumholz, McKee, & Tumlinson 2008b, in press) What is   (  d / R ) (E 0 * / n)? Dust opacity  d and H 2 formation rate R both  Z, so  d / R ~ const CNM dominates shielding, so n is the CNM density CNM density set by pressure balance with WNM, and n CNM  E 0 *, with weak Z dependence.    (  d / R ) (E 0 * / n) ~ 1 in all galaxies! Allowed n CNM FGH curves for MW (Wolfire et al. 2003)

9. Reality Check… Matches observed saturation of HI, with higher  HI at low metallicity! Compare model to BIMA SONG (Blitz & Rosolowsky 2006) and HERA / THINGS (Leroy et al. 2008) surveys, with galaxies binned by metallicity HIH2H2 Matches column needed for molecules to appear, with higher  at lower metallicity!

10. Another Application: DLAs (Krumholz, Ellison, Prochaska, McKee, & Tumlinson, 2009, in preparation)

11. What Does this Imply for SF? t dep = 10 t ff t dep = 1000 t ff Depletion time as a function of  H2 for 2 local galaxies (left, Wong & Blitz 2002) and as a function of L HCN for a sample of local and z ~ 2 galaxies (right, Gao & Solomon 2004, Gao et al. 2007) t dep = 10 t ff t dep = 1000 t ff

12. There is a Universal SFR Clouds convert ~1% of their mass to stars per t ff, regardless of density or environment (Krumholz & McKee 2005; Tan, Krumholz, & McKee 2006; Krumholz & Tan 2007)

13. A Remark on GMCs SFR is simply 0.01  M mol / t ff-mol ! In low  galaxies, GMCs all have  ~ 100 M  pc –2 (Bolatto et al. 2008) due to internal regulation (Krumholz, Matzner, & McKee 2006) In high  galaxies,  GMC must be ≥  gal to maintain hydrostatic balance Luminosity (  mass) vs. radius for galactic and extragalactic GMCs (Bolatto et al. 2008)

14. Putting it Together: The Total Gas Star Formation Law (Krumholz, McKee, & Tumlinson, 2009, in preparation) Super-linear from HI  H 2 conversion Linear from universal GMC properties Super-linear from rising GMC density Lines: theory Contours: THINGS, Bigiel et al. 2008 Symbols: literature data compiled by Bigiel et al. 2008

15. Atomic and Molecular Star Formation Laws HI H2H2

16. A Project for the EVLA In low metallicity, low  gas, the SFR is below Kennicutt law prediction. This may be an effect of low H 2 content. The EVLA can check this directly. Density of star-forming DLAs (points) and Kennicutt law prediction (blue line) (Wolfe & Chen 2006) SF law in LSB galaxies (points) and Kennicutt law (line) (Wyder et al. 2009)

17. Summary Star formation is a 2-step process: (1) convert HI in the ISM into H 2, (2) turn H 2 into stars. Making molecules depends on gas , Z Making stars from molecules depends on t ff Either step can be the rate-limiting one, depending on the environment This has important implications for SF and galaxy evolution at high z, low metallicity