The Irradiated and Stirred ISM of Active Galaxies

The Irradiated and Stirred ISM of Active Galaxies Marco Spaans, Rowin Meijerink (Leiden), Frank Israel (Leiden), Edo Loenen (Leiden), Willem Baan (ASTRON), Dominik Schleicher (Leiden/ESO), Ralf Klessen (Heidelberg) Juan Pablo Perez Beaupuits (Groningen)

PDRs: 6 < E < 13. 6 e. V Heating: Photo-electric emission from grains and cosmic rays n Cooling: Fine-structure lines like [OI] 63, 145; [CII] 158 μm and emission by H 2, CO, H 2 O n 10 e. V photon penetrates 0. 5 mag of dust n

XDRs: E > 1 ke. V Heating: X-ray photo-ionization --> fast electrons; H and H 2 vib excitation; UV emission (Ly α, Lyman-Werner) n Cooling: [Fe. II] 1. 26, 1. 64; [OI] 63; [CII] 158; [Si. II] 35 μm; thermal H 2 vib; gas-dust n 1 ke. V photon penetrates 1022 cm-2 of NH n

n n n PDR (left) with n=105 cm-3 and G=103. 5 XDR with n=105 cm-3 and FX = 5. 1 erg s-1 cm-3 Note NH dependence H 2, C+, C, CO, OH, H 2 O: FIR lines of species trace different regions

A comment on AGN: Relative Size PDR/XDR n 107 M๏ BH at 3% Eddington forh G 0=10 and 1 -100 ke. V powerlaw of slope -1 (with 10% L)

MDRs: how about kinetics? Mechanically Dominated Regions n Turbulent dissipation heats the gas, which leads to IR emission n UV only heats cloud surface n Cosmic rays also heat deep inside cloud, but strongly affect HCO+ n E. g. , at T>100 K: HNC + H HCN + H n

Sources of Turbulence YSOs n SNe n Sloshing motions (accretion) n n Assume 1 -10% efficiency through a turbulent cascade -> mechanical heating competes with normal CR heating for SF rates of 10 – 100 Mo/yr

n E. g. , P cygni profiles in Arp 220: 100 km/s outflow (100 pc scale) g

changes in high density tracers normal mechanical temperature increases E. g. , HNC, HCN, HCO+ affected

Sample of ULIRGs n combined PV, SEST and literature n total of 117 sources, but incomplete: n Note: single dish, so integrated properties – low density gas: CO(1 -0) & CO(2 -1) – high density gas: HCN(1 -0), HNC(1 -0), HCO+(1 -0), CN(2 -1), CS(3 -2) – 110 CO(1 -0), but 32 CO(2 -1) – 84 HCN – only 33 have HCN, HNC and HCO+

Relation with LFIR n n relation LFIR – Lmolecule reflects Kennicutt-Schmidt laws: ~ Σgasα , α=1. 4 Krumholz & Thompson (2007): – if ncrit < nave: α ≈ 1. 5 (KS law) – if ncrit > nave: α ≤ 1 – Note: slope in fits = 1/α ΣSFR

A few fits 2 e 3 1 e 4 CO(1 -0) α ~ 1. 4 CO(2 -1) closer to 1 Others α ≤ 1; black squares OH-MM 3 e 6 4 e 5 3 e 6 2 e 7 2 e 5 1 e 6

Relation with LFIR n n Kennicutt-Schmidt laws: ΣSFR ~ Σgasα , α=1. 4 Krumholtz & Thompson (2007): – if ncrit < nave: α ≈ 1. 5 (K-S law) – if ncrit > nave: α ≤ 1 – Note: slope in fits = 1/α n Our data follow the K&T predictions, but can we learn more?

dense gas and go from SF -> SNe

For some ULIRGs, dense gas tracers that correlate with IR may trace more SN than UV exposure, see Loenen et al. (2008)

Lowering the metallicity to 1% Zo: CO no longer dominant molecular gas coolant

Summary In addition to fine structure lines, CO, HCN, HNC, HCO+ lines are good diagnostics to get to SF properties n Turbulence (and cosmic rays) matter! n

so IR response of the ISM may not be tracing star formation directly; [CII] en [CI] lines probe this directly

n n n How about CRs? PDR model with CR rate = 5 x 10 -15 s-1; so SN rate for ~100 M 0/yr Note small changes in C, OH and H 2 O

In fact, CRs can dominate thermodynamics of molecular gas for star formation rates > 100 Mo/yr; think of Arp 220