1. The First Generation of SMBHs and Their Host Galaxies Columbia University AGN and Galaxy Evolution Castel Gandolfo 3-6 October, 2005 Zoltán (  So ł tan) Haiman Mark Dijkstra, Bence Kocsis

2. Outline 1. Theoretical Expectations – chemistry and cooling in cosmological models 2. Growth of the z~6 Quasars – in hierarchical structure formation theories 3. Catching Proto-Galaxies in Assembling Stage – imaging large scale gas infall in Lyα emission 4. Prospects for Future Detections – radio counts, gravity waves

3. Condensations in Hierarchical Cosmology Smallest scales condense first Jeans mass: ~10 4-5 M  log (Mass/ M  ) collapse redshift 0 10 20 30 234567 Consider linear growth of DM perturbations in concordance cosmology

4. Modeling the Gas in the First Halos: Cooling and Chemistry Collapsed Dark Matter halos virialize, gas shock heated to virial temperature necessary Efficient cooling is a necessary condition for continued contraction following virialization (i.e. for anything “interesting” to happen) Primordial gas chemically simple: H, He, H 2

5. Radiative Cooling Function (H+He gas) cf. Halo virial temperature: log( Temperature / K ) log( cooling rate / erg s -1 cm 3 ) COSMIC TIME MASS SCALE

6. Cooling and Chemistry End of “Dark Age” controlled by abundance of H 2 : abundance of H 2 : (Saslaw & Zipoy 1967 Haiman, Thoul & Loeb 1996) Molecular cooling to T~10 2 K (M~10 5 M  ; z~20) Atomic cooling to T~10 4 K (M~10 8 M  ; z~12) What is the molecular abundance? n H2 /n H ~10 -6 after recombination ~10 -3 in collapsed objects

7. Collapse of Spherical Cloud Haiman, Thoul & Loeb (1996) Clouds with virial temperature T vir ≳ 200 K can form H 2, cool and collapse Gas Phase Chemistry: H + e -  H - +  H - + H  H 2 + e - redshift radius (pc) 10 3 M  10 4 M  10 6 M  10 5 M 

8. 3D Simulations of a Primordial Gas Cloud Abel, Bryan & Norman (AMR) Bromm, Coppi & Larson (SPH)  M tot  10 6 M  z  20 T  200 K n  10 4 cm -3  M = few 100 M  f * < 1 % Temperature Density Gas Phase Chemistry: H + e -  H - +  H - + H  H 2 + e -

9. What forms in these early halos? STARS: FIRST GENERATION METAL FREE - massive stars with harder spectra - boost in ionizing photon rate by a factor of ~ 20 - return to “normal” stellar pops at Z ≳ 10 -3.5 Z ⊙ (Tumlinson & Shull 2001 ; Bromm, Kudritzki & Loeb 2001; Schaerer 2002) SEED BLACK HOLES: (~10 2-6 M ⊙ ) - boost by ~10 in number of ionizing photons/baryon - harder spectra up to hard X-rays - must eventually evolve to quasars and remnant holes [to z~6 super-massive BHs; probed by gravity waves] (Oh; Venkatesan & Shull; Haiman, Abel & Rees; Haiman & Menou)

10. Remnants of Massive Stars Heger et al. 2003 (for single, non-rotating stars) 10M  25M  40M  140M  260M  Z=Z Z=Z  Z=0 metalicity

11. Feedback Processes INTERNAL TO SOURCES - UV flux unbinds gas - supernova expels gas, sweeps up shells - H 2 chemistry (positive and negative) - metals enhance cooling - depends strongly on IMF GLOBAL - H 2 chemistry (positive and negative) - photo-evaporation (minihalos with  <10 km/s) - photo-heating (halos with 10<  <50 km/s) - global dispersion of metals (pop III  pop II) - mechanical (SN blast waves) Do most minihalos fail to form stars or black holes?

12. Global Feedback I. Radiative Soft UV background: H 2 dissociated by 11.2-13.6 eV photons: H 2 +   H 2 (*)  H+H+  ’ this background inevitable and it destroys molecules ~ 1 keV photons promote free electrons  more H 2 H+   H + +e - +  ’ H + e -  H - +  H - + H  H 2 + e - Soft X-ray background: this background from quasars promotes molecule formation ⊝⊕ Haiman, Abel & Rees (2000)

13. Global Feedback II. Entropy Floor? Oh & Haiman 2003; Kuhlen & Madau 2005 Most of 1 st generation objects may form in more massive halos Redshift Normalized Entropy First star creates ~ 100 kpc ionized bubble Star dies after ~10 6 yrs and HII region recombines Fossil HII region cools off CMB T~300 K implies excess entropy Contraction, H 2 formation prevented Depends on density at illumination

14. What happens in T vir >10 4 K halos? These halos, as opposed to minihalos, may be the dominant hosts of the “first generation” of black holes. Behavior of gas has not been studied in same detail as for minihalos (no 3D simulations).

15. What happens in T vir >10 4 K halos? Isothermal atomic cooling at T=10 4 K: Jeans mass remains high Oh & Haiman 2001 most disks would be stable Oh & Haiman 2001 direct collapse to 10 6 M  BH ?  need H 2 to fragment / form stars. cf ~100 M  for stars Bromm & Loeb 2003 Volonteri & Rees 2005

16. Can H 2 form in non-equilibrium chemistry? log( Temperature / K ) log( cooling rate / erg s -1 cm 3 ) universal ratio of n(H 2 )/n(H)~10 -3 independent of density, temperature, background flux Key: gas cools faster than it recombines, leaving extra electrons Oh & Haiman (2002)

17. Conclusion: T vir >10 4 K halos cool to ~100K (Oh & Haiman 2001; Haiman, Spaans & Quataert 2001) Similar to minihalos: Rely on H 2 cooling and fragment on similar (few 100 M  ) scales ? Main difference: Main difference: contract to higher densities less susceptible to feedback HD reduces temperature and fragmentation scale? N.B.: cooling radiation may be observable as Ly  ‘fuzz’ Uehara & Inutsuka 2000 Machida et al. 2005 Johnson & Bromm 2005

18. Outline of Talk 1. Theoretical Expectations – chemistry and cooling in cosmological models 2. Growth of the z~6 Quasars – in hierarchical structure formation theories 3. Catching Proto-Galaxies in Assembling Stage – imaging large scale gas infall in Ly α 4. Predictions for the Future – radio counts, gravity waves

19. High-z Supermassive BHs Example: SDSS 1114-5251 (Fan et al. 2003) z=6.43 M bh  4 x 10 9 M  e-folding (Edd) time: 4 x (  /0.1)  -1 10 7 yr Age of universe (z=6.43) 8 x 10 8 yr  How did this SMBH grow so massive? No. e-foldings needed ln(M bh /M seed ) ~ 20 M seed ~100 M 

20. Growth of High-z Supermassive BHs z=6.43 z=20 CDM merger tree 1. Most of the BH mass from z~15 seeds: must start early! 2. High efficiency (  ≳ 0.2) ruled out, unless seeds very massive 3. A super-Eddington accretion phase is required  min = 10 km/s

21. Assembly history of z=6.43 SDSS quasar M bh =4.6×10 9 M ⊙ M halo =10 13 M ⊙ Lacey & Cole (1994) Luminosity (Eddington): Abundance:  =0.1  =1  min = 30 km/s Assume: Sum of smaller BHs, each growing exponentially from a stellar seed redshift (Haiman 2004) Gravity wave ‘Kick’ Gravity wave ‘Kick’ of > 100 km/s of > 100 km/s Favata et al. (2004) Merritt et al. (2004)

22. Gravitational Lensing of SDSS QSOs Expected Lensing Probability at z=6 ntrinsic lensing probability small,  ∼ 10 -3 - intrinsic lensing probability small,  ∼ 10 -3 (Comeford, Haiman & but magnification bias can boost it to  ∼ 1 - but magnification bias can boost it to  ∼ 1 Schaye 2003) Search for High-Magnification Lensing - for spherical lens, μ  2 produces multiple images - No 2 nd image on HST images of 4 high-z QSOs to 0.3” resolution (Richards et al. 2004) Magnification without Multiple Images (Keeton, Kuhlen & - ellipticity and/or shear can give high μ Haiman 2004) - average over realistic e, γ distributions - dwarfs (NFW), galaxies (SIS), clusters (NFW) Fraction of Lens Systems without a detectable 2 nd image - single image: 5-10% (mostly NFW) - 2 nd image too faint or unresolved: 24-1% (mostly SIS)

23. Outline of Talk 1. Theoretical Expectations – chemistry and cooling in cosmological models 2. Growth of the z~6 Quasars – in hierarchical structure formation theories 3. Catching Proto-Galaxies in Assembling Stage – imaging large scale gas infall in Ly α 4. Predictions for the Future – radio counts, gravity waves

24. Processed ionizing radiation produced by embedded AGN (or other ionizing source) Extended Lyman  Emission Does the AGN turn or while there is still significant infall of material from large (several 100 kpc) scales? Cooling radiation from contracting gas Ly  photons scattered from the nucleus Can we detect contracting extended (size ~R vir ) hydrogen envelope?

25. Isothermal atomic cooling ? log( Temperature / K ) log( cooling rate / erg s -1 cm 3 ) ~ ~ R vir R vir 10 s R vir Birnboim & Dekel (2003); Maller & Bullock (2004)

26. Can We Detect the Cooling Radiation? 2. How many halos can we hope to detect ? ~1 halo per JWST field at z=7 v circ ≥ 150 km/s ~10 halos per field at z=3 blind search in R=5 filter (narrow filter can go deeper) 1. How does cooling halo look like? flux spread over R vir : JWST limiting line flux at 3 < z < 8: “cooling flow” - extended Ly  “blob” 3. What if a bright quasar turns on in a collapsing halo? 10-100 times brighter “fuzz” (Haiman & Rees 2001) (Haiman, Spaans & Quataert 2000) mostly Ly  emission: constrains galaxy formation

27. Lyman  Fuzz Around Young AGN Haiman & Rees 2001 log(T vir /K) log(Flux/erg s -1 cm -2 asec -2 Surface brightness should be detectable Weidinger et al. 2005 Bunker et al. 2004 Bergeron et al. 2002 Steidel et al. 2001 Matsuda et al. 2005 Challenge: interpretation of any possible detection of spatially resolved Lyα emission

28. Ly  Transfer Basics Photons undergo random walk in space+frequency. Different frequency translates to different m.f.p. Moderate optical depth: photon escapes in wing in single flight. Extreme optical depth: photon escapes in single “excursion”. τ ≲ 10 3 τ ≳ 10 3 Less sensitive to profilesMore sensitive to profiles

29. Spectrum Emerging from Static Sphere RedBlue CoreWing

30. Gas Infall vs. Outflow (moderate opt. depth) CF: Spatially resolved fuzz around z~4 quasar (Weidinger et al. 2004) Monte Carlo Model assumes: Dijkstra et al. 2005 gas in NFW halo power-law v(r) central ionizing source (quasar)

31. Effect of Scattering in IGM Transmission Through perturbed IGM ρ(r), v(r) around DM halo (Barkana 2004) Ionizing QSO Impact parameter Characteristic transmission profile extending to red side of Lyα line

32. Gas Infall vs. Outflow CF: Spatially resolved spectrum of Steidel’s LAB # 2 (Wilman et al. 2005) interpretation: Ly α generated by a buried source absorption by 100 kpc shell, swept-up by super-wind Alternative: IGM infall onto a density peak

33. Gas Infall vs. Outflow (Dijkstra et al.) IGM infall (Dijkstra et al.) (Wilman et al.) Absorbing Shell (Wilman et al.)

34. RedBlue -1300 km/s650 km/s Ly  radiation emerges blue-shifted, smaller red peak IGM opacity can make it hard to detect at high z Spectrum Emerging from Collapsing Sphere Dijkstra, Haiman & Spaans, in preparation

35. Diagnostic of Gas Infall: Brightness Profile shallow v(r)steep v(r) Surfacebrightnessprofile radius (arcsec) Blue photons come preferentially from central regions Surface brightness profiles flat (log. slope of -0.5) Scattering vs. Intrinsic effect distinguished using Hα

36. Outline of Talk 1. Theoretical Expectations – chemistry and cooling in cosmological models 2. Growth of the z~6 Quasars – in hierarchical structure formation theories 3. Catching Proto-Galaxies in Assembling Stage – imaging large scale gas infall in Ly α 4. Predictions for the Future – radio counts, gravity waves

37. Direct Detections in Radio Haiman, Quataert & Bower (2004) Model assumes M bh  M halo 5/3 (1+z) (feedback; (feedback; Silk & Rees 1998) Silk & Rees 1998) RL distribution from FIRST-SDSS sample (Ivezic et al. 2003) Duty cycle of 2  10 7 yr Minimum BH mass M bh >10 7 M ⊙ ? M bh >10 7 M ⊙ ?

38. Gravity Waves from BH-BH Mergers (LISA) Menou, Haiman & Narayanan (2001); Volonteri et al. (2004) Tens of mergers per year detectable in LISA frequency band. Can measure Eddington ratio if quasar counterpart is found (possible only to z~1-2) Many other motivations Kocsis, Frei, Haiman & Menou (2005)

39. Can We Identify a Unique Counterpart? Kocsis, Frei, Haiman & Menou (2005) Angular and Radial localization from GW signal alone depends on physical and orbital parameters and orientation Angular Error: large, and dominated by LISA uncertainty Radial Errors: - LISA d L (z) measurement - Cosmological Model  z ≲ 0.005 - Cosmological Model  z ≲ 0.005 - Peculiar velocity - Peculiar velocity - Lensing-induced variations in d L (z):  z  0.03 - Lensing-induced variations in d L (z):  z  0.03 at z=1 at z=1 }

40. Number of Quasars in 3D LISA Error Box Extrapolate known optical QSO LF to M BH ≲ 3x10 7 M ⊙ Assume L/L(edd) ~ 0.3, consistent with recent obs+models Compute mean number in error box (20% lensing correction) Feasible at z<1 for 4x10 5 M ⊙ ≲ M BH ≲ 10 7 M ⊙ Can be extended to z=3 if BHs spin rapidly

41. Conclusions 1. Hosts of the first generation of BHs: ~10 8 M ⊙ dark matter halos collapsing at z~10 (as opposed to minihalos relying on H2-cooling) 2. z~6 QSOs are not strongly lensed. Assembling ≳ 10 9 M ⊙ BHs requires seeds growing uninterrupted since z~15 (and also a super-Eddington growth phase) 3. Lyα halos may offer diagnostic of early stages of the thick collapsing gaseous envelopes around proto-galaxies 4. LISA can measure precise L/L Edd if QSOs accompany GWs

42. Le Fin