Black hole growth and the Eddington limit - Marta Volonteri F. Pacucci, A. Ferrara G. Dubus, J. Silk
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Black hole growth and the Eddington limit Marta Volonteri Institut d’Astrophysique de Paris F. Pacucci, A. Ferrara G. Dubus, J. Silk
High-redshift quasars Quasars have been detected at very large distances, corresponding to a very young age of the Universe. Fan 2012
High-redshift quasars blue/red/green/purple (optical): z>6 orange (X-ray) High luminosity and large estimated MBH masses Some fainter sources found in X-ray (Fiore+12, Giallongo+15) Wu et al. 2015 As massive as the largest MBHs today, but when the Universe was ~ Gyr old! Gultekin et al. 2009
The Eddington limit: the outward radiation pressure equals to the inward gravitational force If arad>|g| radiation pushes away the gas, and further accretion is halted Note: spherical symmetry
The Eddington limit: the outward radiation pressure equals to the inward gravitational force ε: radiative efficiency, ~0.05-0.3 If H only: Ledd=1.3e38 erg/s (M/Msun)
For a BH accreting at a fraction fEdd of the Eddington limit, mass grows in time as: 1−ε t ( fEdd ) M (t) = M in e 0.45Gyr SDSS z=6 quasars ULASJ112010641 ULAS J1120 @ z=7.1 M=2x109 Msun t~0.75 Gyr ε~0.1 M in>300-ish Msun and fEdd~0.5-1
Minimum Eddington ratio that a MBH must sustain continuously throughout the whole cosmic history to reach a given mass by a given redshift tEdd ε " M (z) % fEdd,min = ln $ ' t H (z) 1− ε # M 0 & E.g., to reach 1e9 msun by z=4 a BH with initial mass 1e4 msun has to accrete at fEdd=0.4 for the whole time or at fEdd=1 for 40% of the time
E.g., to reach 1e10 msun by z=4 a BH with initial mass 1e4 msun has to accrete at fEdd=0.55 for the whole time or at fEdd=1 for 55% of the time
Eddington ratio distribution peaks near fEdd ∼ 0.05 1 Kelly & Shen
The feeding of high-z MBHs - Estimate Eddington rate for BHs in Horizon-AGN -- 3x106 Mpc3 (Dubois+14) - Supercritical inflows possible, ~10% at z>6 - What happens when they reach the MBH? Volonteri, Silk & Dubus 2015
Super-Eddington accretion? Super-Eddington accretion vs super-Eddington luminosity Highly super-Eddington accretion does not imply highly super-Eddington luminosities Low “effective” radiative efficiency: ε
• Mass supply rate exceeds Eddington limit: M! >> M! Edd • Energy released GMM! c ~ LE where τ = v r c • Energy accumulates at r
Accretion discs If gas has low angular momentum => gas does not fall radially on the BH, and forms an accretion disc capture radius: where gas becomes bound to the BH accretion disc radius: set by gas angular momentum
Accretion disks • Thin disk – ~ all dissipated energy radiated away – ~ circular Keplerian orbits, vertical structure decouples – energy transport: internal torque • Slim disk – gas retains enough pressure to affect radial balance – energy transport: torque + advection • Radiatively inefficient disk – Due to low density or high optical depth (Eddington limit) – must dispose of extra energy, mass, or angular momentum to avoid becoming unbound Courtesy of M. Begelman
Starlike accretion • Dynamical conditions don’t allow a bound disk- like flow if if specific angular momentum is too small compared to Keplerian ZEro BeRnoulli Accretion Flow (Coughlin & Begelman 2013) Courtesy of M. Begelman
GRMHD simulations https://www.youtube.com/watch? v=q2DeKxUHae4 McKinney et al. 2014
GRMHD simulations Sadowski et al. 2014
GRMHD simulations Sadowski et al. 2014
Trapping all of it Trapping of radiation: the time for photons to escape the disk exceeds the timescale for accretion Trapped photons are advected inward with the gas, rather than diffuse out ! M! $ rtrap = 10 # & rg " M! Edd % L ! M! $ Luminosity highly suppressed ~ ln # & LEdd " M! Edd %
Trapping all of it If gas has low angular momentum => the whole accretion disc may be within the trapping radius and luminosity highly suppressed capture radius: where gas becomes bound to the BH accretion disc radius: set by gas angular momentum trapping radius: set by inflow rate
Trapping all of it • Rg is the capture radius, aka Bondi radius: • RD is the disc radius defined by the angular momentum barrier, with λ parameterizing angular momentum loss: ! M! $ GM ! M! $ G • impose RD< rtrap = 10 # ! & 2 = 10 # & 2 " M Edd % c " tEdd % c Volonteri, Silk & Dubus 2015
Trapping all of it ! M! $ GM rtrap = 10 # & 2 ! " M Edd % c
Trapping all of it 105 Msun MBH could grow to ∼108 Msun, in ∼106 years => boost of ~3×102 vs Eddington - gas inflow rate: 1-10 Msun/yr (∼1% of the free fall rate) - only gas with low angular momentum (λ~1% of the mean) is accreted Only short periods needed to ease constraints (e.g.Volonteri & Rees 2005,Volonteri, Silk & Dubus 2015)
Trapping all of it Trapping condition: The lower the gas angular momentum, via λ, the longer accretion can continue, the higher the final MBH mass. In galaxies with much low-angular momentum gas near the center the MBH can get to a higher mass at fixed gas velocity dispersion. Volonteri, Silk, Dubus 2015
Inflow rates and column density X-ray absorption: photoelectric absorption τ=σNH Compton-Thin: 21
Inflow rates and column density • Free fall rate • Given the inverse dependence on the MBH mass, there is a mass above which the opacity through the sphere < 1 and the MBH is unveiled Volonteri, Silk, Dubus 2015
Inflow rates and column density The gas density is very high and so accretion is obscured, regardless of any assumption on the accretion disc structure and properties - ULASJ1120: L∼ 2×1047 erg/s, MBH∼ 2×109M⊙, σ∼ 100 km/s - J1148: L∼ 7 × 1047 erg/s, MBH ∼ 6 × 109 M⊙, σ∼ 160 km/s => fff∼ 0.15 − 0.35 => NH ∼ few×1022cm−2 (NH < 1023 cm−2 , Moretti et al. 2014).
How do the first black holes grow? Mgas=107 Msun • Spherical symmetry • Radiation-hydrodynamic simulation • Accretion disc is unresolved • No magnetic fields • Cooling: bremsstrahlung and atomic 105 Msun • Opacity: Free-Free and Bound-Free • Gas density profile extracted from cosmological simulations of direct collapse BH formation (Latif+2014) • Standard thin accretion disc • Slim disc (supercritical accretion) domain: 10−3 pc to 20 pc Pacucci & Ferrara 2015, Pacucci,Volonteri & Ferrara 2015
How do the first black holes grow? • Standard thin accretion disc • M! m= M Edd • Slim disc (supercritical accretion) Pacucci,Volonteri & Ferrara 2015
How do the first black holes grow? duty cycle • • M / M Edd duty cycle Standard accretion: L \propto Mdot Luminosity mildly super-Eddington Pacucci,Volonteri & Ferrara 2015
How do the first black holes grow?
How do the first black holes grow? duty cycle • • M / M Edd duty cycle Slim disc accretion: L \propto ln(Mdot) Luminosity sub-Eddington, while accretion super-critical Pacucci et al. 2015
How do the first black holes grow? Pacucci,Volonteri & Ferrara 2015
How do the first black holes grow? The timescale for outflows to develop is ~1/ε ~ 60 times longer in radiatively inefficienf flows than in the standard accretion scenario The BH can consume most of the available gas before feedback becomes important Pacucci,Volonteri & Ferrara 2015
Summary – growing MBHs • Highly super-Eddington accretion does not imply highly super-Eddington luminosities • Radiatively inefficient accretion flows may ensure a continuous growth with accretion rates largely exceeding the Eddington value, but with sub-Eddington luminosities
You can also read