The Interstellar Medium Fall 2014 Lecturer: Dr. Paul van der Werf Oortgebouw 565, ext 5883 pvdwerf@strw.leidenuniv.nl Assistant: Kirstin Doney Huygenslaboratorium 528 doney@strw.leidenuniv.nl
Class Schedule (full details on website) 1. Introduction. Basic physical processes Draine Ch 1 2. Emission and absorption processes. Radiative transfer Draine Ch 6 & 7 3. The HI 21cm line Draine Ch 8 4. Ionization and recombination Draine Ch 12, 13 & parts of Ch 14 5. Photoionization and HII regions Draine Ch 15 (parts) 6. Collisional excitation. Nebular diagnostics Draine Ch 17 & parts of Ch 18 7. Molecular energy levels and excitation and radiative trapping Draine Ch 5 (parts) & 19 8. Radiative trapping and molecular clouds Draine Ch 19 & 32 (parts) 9. Interstellar dust Draine Ch 21 & parts of 23 & 24 10. Thermal balance and the two-phase model of the ISM Draine Ch 27 (parts), 29 & 30 11. Shocks and the 3-phase ISM Parts of Draine Ch 35, 36, & 39 12. PDRs, XDRs, & extragalactic ISM Parts of Draine Ch 31, 36, 41 & 42
Molecular clouds H 2 formation and destruction CO formation and destruction Photon-dominated regions (PDRs) Heating and cooling Shocks in molecular clouds Star formation on Galactic scales ISM in (ultra)luminous infrared galaxies
H 2 photodissociation In general, photo-dissociation is a quite simple process: absorption into repulsive excited state leads to dissociation: direct photo-dissociation See H 2+ potential curves for example For H 2, however, there are no allowed transitions to repulsive states from the ground state at energies less than 13.6 ev no direct photodissociation possible However, photodissociation of H 2 can occur via an indirect, 2-step process: spontaneous radiative dissociation
Potential curves for H 2+ ion R H-----H H + H + D e R e
H 2 photodissociation Direct photodissociation from ground state forbidden 87% of absorptions into B and C states are followed by emission back into bound vibrational levels of the X state 13% of the absorptions are followed by emission into the unbound vibrational continuum, leading to dissociation (as indicated in the figure)
H 2 formation on grains 3+4 5.
Diffusive mechanism X+g:Y X:g:Y X Y:g XY+g sticking diffusion + desorption molecule formation Process proceeds in several steps: 1. H atom must collide with grain: H + g 2. Colliding atom must stick to surface: H + g H : g 3. H atom must be retained until another atom gets absorbed: H : g H : g : H 4. H atoms must be mobile to find each other and form bond: H : g : H H H : g 5. H 2 must be ejected from surface: H H : g H 2 + g
Self-shielding of H 2 H 2 UV absorption lines leading to dissociation become very optically thick at small depth into the cloud; this is called self-shielding H 2 molecules at edge of cloud absorb all available photons at certain wavelengths, so that molecules lying deeper in the cloud see virtually no photons at all, and are not dissociated
CO formation and destruction CO is good tracer of H 2 (although not without problems) But: at edge of cloud, most of carbon is C + Transition C + C CO with increasing depth CO is very stable (D e = 11.09 ev 1118 Å) can only be dissociated at 912 Å < λ < 1118 Å
CO photodissociation Like H 2, CO has no direct dissociation channels dissociation through line absorption self-shielding, but at greater depth than H 2 because of smaller abundance At A V 1 2 mag, CO / H 2 increases from 10 7 to 10 4
Self-shielding of CO and H 2 Photodissociation rates - Note that H 2 lines can shield CO UV lines: mutual shielding
Densities of major species in cloud envelope T=15 K n H =500 cm -3 I UV =1 Edge
Photon-dominated regions (PDRs) Cloud surfaces are examples of PDRs, i.e., clouds in which UV photons control the physical and chemical state of the cloud Traditionally, PDRs are dense molecular clouds located close to an OB star, in which the UV radiation field is enhanced by a factor of 10 5 w.r.t. average interstellar radiation field Example: Orion Bar PDRs show very strong atomic fine-structure lines e.g. [C II] 158 μm, [C I] 610 μm, [O I] 63 μm And submillimetre lines of molecules e.g. CO mid-j lines (4-3 to 9-8: warm, dense molecular gas)
PDR structure
Orion Bar PDR Note layered structure! Tielens et al. 1993
Thermal balance in molecular clouds Cloud envelope: heating by photoelectric effect from UV radiation cooling by fine-structure lines ([CII] 158 µm & [OI] 63 µm) Cloud core: heating mostly by cosmic rays (but see later for mechanical heating!) cooling dominated by molecular lines, in particular CO
Cooling in molecular gas as function of T and n
Shocks in molecular clouds: MHD effects So far considered only single-fluid shocks Normally interstellar gas consists of 3 fluids: (i) neutral particles (ii) ions (iii) electrons Fluids can develop different flow velocities and temperatures For B 0 information travels by MHD waves Perpendicular to B the propagation speed is the magnetosonic speed
Magnetic Precursors: C-shocks Alfvén speed for decoupled ion-electron fluid can be much larger than Alfvén speed for coupled neutral-ion-electron fluid In many cases: C S <v A,n < v s < v A,ie Ion-electron plasma sends information ahead of disturbance to inform pre-shock plasma that compression is coming: magnetic precursor Compression is sub-sonic and transition is smooth and continuous C-shock Ions couple by collisions to neutrals
Schematic structure of J- and C-shocks
Temperatures in C-shocks Draine & Katz 1986
Post-shock temperature structure of fast J-shock Three regions: Hot, UV production Recombination, Lyα absorption Molecule formation Grains are weakly coupled to gas, so that T gr T gas
Comparison of J-shocks and C-shocks J ( jump )-shocks: v S 50 km/s; fractional ionization high Shock abrupt Neutrals and ions tied into single fluid T high: T 40 (v S /km s -1 ) 2 Molecules destroyed (but can reform in cooling post-shock gas) Most of radiation in ultraviolet C ( continuous )-shocks: v S 50 km/s; fractional ionization low Gas variables (T, ρ, v) change continuously Ions ahead of neutrals; drag modifies neutral flow T i T n ; both much lower than in J-shocks Gas heated but molecules (mostly) not destroyed! Most of radiation in infrared
Star formation efficiency Free fall time Milky Way cloud mass ~4 10 9 M Theoretical star formation rate Mcloud/tff ~ 1400 M /yr Observed star formation rate across the MW ~ 3 M /yr cf. starburst galaxy: tens to hundreds M /yr Observed star formation rate inside star forming regions between 3 6% (Taurus, Ophiuchus,...) and 30-40% (Orion) Explanations Clouds are supported against free collapse (?) Star formation is inefficient in turning mass into stars (?)
(Ultra)luminous infrared galaxies or (U)LIRGs ULIRGs : Milky Way : Galactic GMCs : OMC -1: Orion BN - KL : L M FIR H 2 100 L 1.5 L 1.8 L 54 L 400 L M L IR /L CO SFR/M H 2 1 SFE M M M 1 1 1 M 1 (Gao & Solomon 2001) L IR SFR
Power-source of (U)LIRGs If extinction is moderate: starburst and AGN can be distinguished using optical/ir spectra (fine-structure lines revealing the hardness of the radiation field) If extinction high: need to go to submm; mostly molecular lines (U)LIRGs were much more common at high redshift and form and important mode of galaxy evolution What powers them: extreme starburst or AGN ( babies vs. monsters discussion)?
Infrared spectrum of the starburst galaxy M 82 strong PAHs low ionization stages
Infrared spectrum of Circinus (AGN & Starburst) weak PAHs high ionization stages
X-ray dominated regions (XDRs) Differences between PDRs & XDRs: X-rays penetrate much larger column densities than UV photons Gas heating efficiency in XDRs is 10 50%, compared to <3% in PDRs Dust heating much more efficient in PDRs than in XDRs High ionization levels in XDRs drive chemistry PDRs: star formation; XDR: active galactic nucleus
PDRs vs. XDRs: CO lines XDRs produce larger column densities of warmer gas Identical incident energy densities give very different CO spectra Very high J CO lines are excellent XDR tracers (Spaans & Meijerink 2008)
Mrk231 (nearest QSO) submm spectrum (Van der Werf et al., 2010)
CO excitation in Mrk231: PDR+XDR 2 PDRs + XDR 6.4:1:4.0 High-J lines require XDR n=10 4.2, F X =28 * n=10 3.5, G 0 =10 2.0 n=10 5.0, G 0 =10 3.5 Low-J lines powered by star formation. * 28 erg cm -2 s -1 G 0 =10 4.2
General results submm spectra of (U)LIRGs AGNs reveal themselves through high-j CO line emission; mid-j CO line emission powered by star formation Surprise 1: in star-forming galaxies more warm molecular gas than can be explained by UV radiation (PDRs) most likely: dissipation of turbulence as an important heating mechanism Surprise 2: massive molecular outflows removing the fuel for star formation
Outflow tracers: P Cygni profiles OH, OH +, CH +, H 2 O +, HF, H 2 O and NH 3 lines reveal massive molecular outflows
Blowout into the halo ( superwind ) NGC 4631 optical Red = Ha blue = X-ray
Hot topics & questions Balance of gas inflow and outflow as regulating principle for star formation in galaxies What drives outflows? Supernovae? AGN? Radiation pressure? Surprising role for dissipation of turbulence in heating molecular gas Many key issues still ununderstood, e.g.: Relation starburst-agn Star formation law(s) Origin and universality of stellar Initial Mass Function Star formation in the first galaxies &c, &c, &c