Supermassive Black Hole Formation in Galactic Nuclei

Supermassive Black Hole Formation in Galactic Nuclei Melvyn B. Davies Department of Astronomy and Theoretical Physics Lund University Ross Church (Lund), Cole Miller (Maryland), Serge Nzoke (Lund), Jillian Bellovary (Michigan)

Today s seminar 1. Formation of SMBHs at high redshift 2. Absence of SMBHs in local galaxies 3. The centre of our own galaxy With nuclear stellar clusters being a common theme.

Nuclear stellar clusters: Found in a large fraction of galactic nuclei. Typically 10-100 times more massive than globular clusters, and slightly larger. (eg Carollo et al 1997, 1998; Böker et al 2002; Walcher et al 2005; Côté et al 2006)

Part 1: SMBHs at high Z

A Challenge: Mortlock et al A luminous quasar at a redshift of z = 7.085, 2011, Nature, 474, 616 This is a SMBH made very early on (770 Myr). SMBH mass is 2 billion solar masses How can SMBHs form and grow so quickly (note that Eddington limited growth doubling time is around 30 Myr)? Need to make a 10 5 solar mass seed black hole at around z = 11 (cf Di Matteo et al 2012).

Literature includes: Gas in-flow into galactic nuclei and first quasars (Mayer et al 2010; Di Matteo et al 2012). Quasistars (Dotan, Rossi & Shaviv 2011). General reviews (Alexander & Hickox 2012; Volonteri & Bellovary 2012). Cluster evolution (Quinlan & Shapiro 1990).

A Model Nuclear Stellar Cluster Halo Core

Two processes: Energy loss via two-body scattering from cluster core leading to core collapse. Heating via binary-single encounters which could prevent or delay core collapse. If binaries are tight enough they will spiral together and merge before they can heat the cluster.

Timescales for encounters and inspiral τ 2+1 6 10 8 x M 2 c,6 v 3,10 where yr a bin 1000R x m bh v 2,10 τ gr 5 10 19 m bh v 8,10 x 4 1 e 2 7/2 yr (Davies, Miller & Bellovary 2011)

Binary-single encounter and GR inspiral timescales (Davies, Miller & Bellovary 2011)

KEY IDEAS: Addition of gas into nuclear stellar cluster leads to significant contraction in core and increase in cluster velocity dispersion (cf Mayer et al 2010). Binaries can no longer support cluster which undergoes core collapse. Tight binaries merge but are retained (due to high velocity dispersion) to go on to merge with other objects thus building up a seed SMBH.

BOTTOM LINE #1: SMBH will reach a mass of around 10 5 solar masses from stellar-mass BHs, NSs, and WDs within cluster. SMBH formed within roughly 100 Myr of gas infall into cluster (ie z around 11 or 12). Eddington-limited growth onto moderately spinning black hole would see growth to ~10 9 solar masses by z ~7. (Davies, Miller & Bellovary 2011)

Part 2: No SMBHs in some local galaxies

M33 Why no SMBH, even today? (but there is a nuclear stellar cluster)

KEY IDEA: Above a critical velocity dispersion (around 40 km/s), massive BHs may form in relaxed stellar systems. This is because primordial binaries cannot support the cluster against core collapse.

Paths to SMBH formation in a stellar cluster (Miller & Davies 2012)

BOTTOM LINE #2: M33-like nuclear stellar clusters have low velocity dispersions and so might avoid SMBH formation. (Miller & Davies 2012)

Part 3: The centre of our galaxy

The Galactic Centre (Genzel et al 2003)

S-Star Orbits 1 arcsec = 0.04 pc From stellar orbits, we know mass of central object is about 4.3 million solar masses. (Gillessen et al 2009)

The Galactic Centre Contains a Disc of Young Stars A flatter IMF has been suggested for the young stellar disc (Bartko et al 2010) A young stellar disc made every 50-100 Myr could produce a large fraction of all stars in the inner pc. (Paumard et al 2006)

Observations: early vs late-type stars (Buchholz et al 2009)

Ways to remove red-giant population: 1) Stellar collisions RG-BH (clobber giant) MS-MS (change range of stellar masses) MS-CO (destroy stars before giant phase) 2) Star-disc interactions Strip RG as it ascends giant branch 3) Illumination from active SMBH Accretion onto SMBH causes illumination of central stellar population

RG-BH collisions RG-BH collision with v =800kms -1,Rmin=10R, 1M giant, 10M BH (Dale et al 2009)

(Armitage, Davies & Zurek 1996)

Fermi Bubble (Su, Slatyer & Finkbeiner 2010)

Effects of Stellar Collisions We performed a series of Monte-Carlo calculations building up the stellar population over time, whilst allowing for coagulations and destructive collisions. Produce 10 Msolar BH for Produce 1.4 Msolar NS for M 20M 20M M 8M Otherwise produce a 0.6 Msolar WD (Nzoke et al in prep)

Mass enclosed within 0.4 pc (in kmsolar) A B C MS 304 401 109 RG 3 2 3 WD 126 438 388 NS 19 62 504 BH 34 103 2483 Total 486 1006 3487 Schödel et al (2007) number about 800 kmsolar

Effects of Star-Disc Interactions Modelled the evolution of a population of stars allowing for stripping when passing through a gaseous disc. Then observe cluster of stars to see how many late-type stars are visible. (Davies & Church, in prep)

Evolution of a solar-like star 10 1M 12 mk 14 16 0 10 8 2 10 8 3 10 8 (t 12.15 Gyr)/yr

Effects of Star-Disc Interactions 10 10 11 11 12 12 mk 13 mk 13 14 14 15 15 16 0.6 1 2 3 6 R 2d /arcsec 16 0.6 1 2 3 6 R 2d /arcsec RGB RC/HB AGB RGB RC/HB AGB (Davies & Church, in prep)

BOTTOM LINE #3: If disc-mode star formation is top-heavy and common, the galactic centre could contain a very large population of stellar-mass BHs. Disc-mode star formation could also affect any current stellar population via star-disc interactions and illumination from an active galaxy. Stellar collisions, star-disc interactions, and illumination from an active SMBH could all potentially contribute to the observed depletion of RGs.

Summary Gas inflow into nuclear star clusters may accelerate SMBH formation at high redshift. SMBH formation may be inevitable in nuclear stellar clusters having sufficiently high velocity dispersion. M33-like nuclear stellar clusters have low velocity dispersions and so might avoid SMBH formation. Missing RGs in our own galaxy could be telling us about the past history and linked with periods of nuclear activity and SMBH growth.