Black hole X-ray binaries V: Formation and evolution of black hole - PowerPoint PPT Presentation

Black hole X-ray binaries V: Formation and evolution of black hole binaries Thomas J. Maccarone Jan 14, 2017, 2nd Fudan Winter School on Black Hole Astrophysics

Overview Phenomena associated with binary evolution Formation of close binaries Evolution of close binaries Dynamics of star clusters Special considerations for understanding double black hole binaries

Problems solved by binary evolution With compact objects Best place to understand black hole accretion Type I supernovae Millisecond pulsars and precision tests of GR Gravitational wave sources Short GRBs? Thorne Zytkow objects? r-process elements? Source of reionization of the Universe? Without compact objects Extrinsic S stars, Wing Ford band and the IMF Mergers of massive stars and the IMF Blue stragglers

General properties of binaries ~60% of stars are binary stars, and the number is higher for massive stars (“3 out of ever 2 stars is in a binary” – Payne-Gaposchkin) Preference for equal mass ratios (Sana et al. 2012) Periods log normal, with median P~180 yr (Duquennoy & Mayor 1992) Eccentricities “thermally” distributed for P > P_circ

General properties of binaries Sana et al. 2012: study of Tarantula Nebula

End states of single star evolution From Clausen, Piro & Ott 2015, for solar metallicity, non-rotating stars Heger & Woosley 2002

Black hole-neutron star massgap? Most reasonable intrinsic distributions of masses that match the data have a real gap between NS & BH mass ranges Not yet clear whether there is a selection bias against low mass black holes If established, it suggests that there is a rapid explosion of the star once the core collapse starts Birth mass may not be current mass (e.g. Moreno Mendez 2011; Fragos & McClintock 2015) Ozel et al. 2010; Farr et al. 2011; Belczynski et al. 2012

Natal kicks Arzoumanian et al. 2002

Kicks in binaries: the “Blaauw” kick Blaauw 1961: trying to explain “runaway stars” Supernova leads to mass loss If supernova is in a binary, the mass takes away linear momentum from orbital motions There is a recoil on the binary

Space distributions of X-ray binaries Figure: Jonker & Nelemans 2004 Dots are black holes Scale height distributions are quite similar Suggests that the natal kicks are similar This is difficult for current supernova theory to explain

What would happen... To the Earth if the Sun suddenly lost half its mass in an explosion?

Keeping binaries bound If >1/2 of the binary's mass is lost, it becomes unbound With a highly fine-tuned kick, this can be avoided (Kalogera & Webbink 1998), but when < ½ of the mass is lost, kicks can also unbind binaries Some other process must lead to low mass X-ray binaries (remember also that masses in binaries are usually correlated)

Natal kick versus Blaauw kick Blauuw kick only in the direction of the orbital plane Natal kicks can be in any direction Can look for misalignment of black hole spin Can also maybe eventually get position angles for orbits on plane of sky

Common envelope evolution Cores fall toward one another Gravitational energy is released Envelope is expelled Original idea: Paczynski 1976 Recent review: Ivanova et al. 2013 Red novae may be common envelopes in action Very hard problem to solve from pure theory

Summary of issues to understand formation Initial binary fraction (note: we ignore triples, but they might be important) Initial binary mass ratio distribution Initial orbital period distribution Evolution of binaries (circularization, loss of angular momentum) Standard stellar evolution (core evolution and wind mass loss) Supernova process – natal kicks and initial mass-final mass relation Common envelope evolution process THEN we have all the problems with the later evolution after the compact object forms (non-conservative accretion due to disk winds!!) And some of the problems above come up again if we want to form compact object mergers

Donor star abundances GRO J1655-40: 6-10 times higher in alpha elements than Sun, but not in other elements (Israelian et al. 1999). Possible evidence of supernova pollution V404 Cyg: mild oxygen enhancement (Gonzalez Hernandez et al. 2011) A0620-00: solar, except for strong lithium excess (Gonzalez Hernandez et al. 2004) Cen X-4: neutron star, some Ni, Ti enhancement, mildly supersolar [Fe/H]=0.23, very high lithium (Gonzalez Hernandez et al. 2004), perhaps polluted XTE J1118+480: solar to mildly supersolar (Gonzalez-Hernandez et al. 2006) Lithium preservation due to tidal locking (Maccarone, Jonker & Sills 2004), demonstrated by Li6/Li7 ratio (Casares et al. 2007) Ultracompact X-ray binaries have white dwarf donors. None are dynamically confirmed to have black holes, but there are two strong candidates, both in globular clusters (extragalactic: Maccarone et al. 2007; Zepf et al. 2008; Galactic: Miller-Jones et al. 2014)

Globular clusters are X-ray binary factories Left: 47 Tucanae, figure from APOD by Craig Heinke  Right: NGC 4472, Maccarone, Kundu & Zepf 2003 

Globular clusters are X-ray binary factories II  Milky Way: ~10% of low mass X-ray binaries in globular clusters, but only ~0.1% of stars (e.g. Clark 1975)  Elliptical galaxies: ~50% of low mass X-ray binaries in globular clusters Peacock et al. 2010

Mass segregation Energy exchange happens between stars in star clusters Approach a Maxwellian distribution, like in a gas Heaviest objects have lowest speeds They then sink to the center Massive objects, like black holes, then are preferentially located in center of cluster

Dynamical formation of X-ray binaries  Tidal capture (Fabian et al. 1975)  medium period systems  Exchange encounter (Hills 1976)  Long period systems  Direct collision (Verbunt 1987)  Systems with period less than ½ hour Figure from Funato et al. 2007

Why mass transfer takes place  Gas pressure at inner Lagrange point  Pressure scale height of star is ~10^-4 binary separation, so binary must remain circular

Ways to keep star in Roche lobe contact Transfer from light to heavy object expands the binary if angular momentum is conserved Must either shrink the binary or grow the radius of the donor star to stay in contact Donor star can grow through standard expansion off the main sequence (occurs on “nuclear timescale”)

Ways to keep star in Roche lobe contact II Magnetic braking – donor star has a weak wind and a magnetic field Plasma in wind puts a drag on the star Schatzman 1962 for early work, Knigge, Baraffe & Patterson 2013 for a comprehensive treatment on cataclysmic variables

Ways to keep star in Roche lobe contact III Gravitational radiation can transport angular momentum outside a binary as well Long period: usually nuclear evolution Intermediate period (~3-10 hrs): magnetic braking Short period (<~3 hours): gravitational radiation Ultracompact binaries: have degenerate donors, which expand as they lose mass, and evolve to longer periods even with gravitational radiation keeping them in contact King, Kolb & Burderi 1996 – useful paper for understanding mdot for a given binary set-up

Unstable mass transfer If the donor star is more massive than the accretor, mass transfer generally shrinks the binary This leads to unstable mass transfer If the donor envelope is convective, this occurs on the dynamical (i.e. freeefall) timescale of the donor, and often leads to a common envelope or merger If the envelope is radiative, this occurs on the thermal timescale of the donor star, and often leads to ultraluminous X-ray sources

Making gravitational wave sources Complicated path of binary evolution must be followed There are many ways things can go wrong between starting with two massive stars and ending with a merger of compact objects Figure from Postnov & Yungelson 2014

The initial LIGO discovery Two ~30 solar mass black holes This is extremely difficult to do through normal channels of binary evolution Globular cluster formation channel is quite viable Main alternative is chemically homogeneous evolution in a very close binary

Chemically homogeneous evolution Mandel & de Mink 2016 Rotation → convection → chemically homogeneous star Allows core of star to grow in mass to be nearly the whole star Allows star to collapse to much more massive black hole Also requires a close binary

Conclusions Understanding the evolution of binary stars is essential to a wide range of problems in astrophysics Binary evolution relies on many complicated processes This is a field with a long-term future