arXiv subjects
Publications and source records attributed to Jonathan C. Tan.
Supermassive black hole accretion and star formation appear intimately connected. I review the observational and theoretical evidence for this statement. I then discuss how focussed studies of two systems, our Galactic Center and the nucleus of M87, can help to improve our understanding of these processes.
I review the status of massive star formation theories: accretion from collapsing, massive, turbulent cores; competitive accretion; and stellar collisions. I conclude the observational and theoretical evidence favors the first of these models. I then discuss: the initial conditions of star cluster formation as traced by infrared dark clouds; the cluster formation timescale; and comparison of the initial cluster mass function in different galactic environments.
It has been known for more than 30 years that star formation in giant molecular clouds (GMCs) is slow, in the sense that only ~1% of the gas forms stars every free-fall time. This result is entirely independent of any particular model of molecular cloud lifetime or evolution. Here we survey observational data on higher density objects in the interstellar medium, including infrared dark clouds and dense molecular clumps, to determine if these objects form stars slowly like GMCs, or rapidly, converting a significant fraction of their mass into stars in one free-fall time. We find no evidence for a transition from slow to rapid star formation in structures covering three orders of magnitude in density. This has important implications for models of star formation, since competing models make differing predictions for the characteristic density at which star formation should transition from slow to rapid. The data are inconsistent with models that predict that star clusters form rapidly and in free-fall collapse. Magnetic- and turbulence-regulated star formation models can reproduce the observations qualitatively, and the turbulence-regulated star formation model of Krumholz & McKee quantitatively reproduces the infrared-HCN luminosity correlation recently reported by Gao & Solomon. Slow star formation also implies that the process of star cluster formation cannot be one of global collapse, but must instead proceed over many free-fall times. This suggests that turbulence in star-forming clumps must be driven, and that the competitive accretion mechanism does not operate in typical cluster-forming molecular clumps.
We argue that rich star clusters take at least several local dynamical times to form, and so are quasi-equilibrium structures during their assembly. Observations supporting this conclusion include morphologies of star-forming clumps, momentum flux of protostellar outflows from forming clusters, age spreads of stars in the Orion Nebula Cluster (ONC) and other clusters, and the age of a dynamical ejection event from the ONC. We show that these long formation timescales are consistent with the expected star formation rate in turbulent gas, as recently evaluated by Krumholz & McKee. Finally, we discuss the implications of these timescales for star formation efficiencies, the disruption of gas by stellar feedback, mass segregation of stars, and the longevity of turbulence in molecular clumps.
The luminosities of the centers of nearby elliptical galaxies are very low compared to models of thin disc accretion to their black holes at the Bondi rate, typically a few hundredths to a few tenths of a solar mass per year. This has motivated models of inefficiently-radiated accretion that invoke weak electron-ion thermal coupling, and/or inhibited accretion rates due to convection or outflows. Here we point out that even if such processes are operating, a significant fraction of the accreting gas is prevented from reaching the central black hole because it condenses into stars in a gravitationally unstable disc. Star formation occurs inside the Bondi radius (typically ~100pc in giant ellipticals), but still relatively far from the black hole in terms of Schwarzschild radii. Star formation depletes and heats the gas disc, eventually leading to a marginally stable, but much reduced, accretion flow to the black hole. We predict the presence of cold (~100K), dusty gas discs, containing clustered H-alpha emission and occasional type II supernovae, both resulting from the presence of massive stars. Star formation accounts for several features of the M87 system: a thin disc, traced by H-alpha emission, is observed on scales of about 100pc, with features reminiscent of spiral arms and dust lanes; the star formation rate inferred from the intensity of H-alpha emission is consistent with the Bondi accretion rate of the system. Star formation may therefore help suppress accretion onto the central engines of massive ellipticals. We also discuss some implications for the fueling of the Galactic center and quasars.
I review some important questions in the field of massive star formation: What are the initial conditions for proto star clusters and how do they arise? What are the initial conditions for individual massive star formation within star clusters? How do massive protostars accumulate their mass? I compare the Turbulent Core Model (McKee & Tan 2003) to several nearby regions, including Orion KL. Here I also discuss the origin of BN's high proper motion.
In the present-day universe, it appears that most, and perhaps all, massive stars are born in star clusters. It also appears that all star clusters contain stars drawn from an approximately universal initial mass function, so that almost all rich young star clusters contain massive stars. In this review I discuss the physical processes associated with both massive star formation and with star cluster formation. First I summarize the observed properties of star-forming gas clumps, then address the following questions. How do these clumps emerge from giant molecular clouds? In these clustered environments, how do individual stars form and gain mass? Can a forming star cluster be treated as an equilibrium system or is this process too rapid for equilibrium to be established? How does feedback affect the formation process?
We consider whether the variability properties of Gamma-Ray Bursts (GRBs) that produce bright optical and longer wavelength transient afterglows (A-GRBs) are the same as a larger, inclusive sample of bright, long-duration GRBs, selected only by their gamma-ray emission. This sample may include a significant population of physically distinct ``dark'' or ``faint-afterglow'' GRBs with different variability properties, or may be composed of a single population, some of which lack afterglows only because of observational selection effects. We argue that the structure function is the most appropriate method for measuring the variability of bursts because of their transient and aperiodic nature. We define a simple statistic: the ratio of the integrated structure function from 0.1 to 1 s compared to that from 0.1 to 10 s, as measured in the observer frame. To avoid instrumental effects we restrict our analysis to GRBs with BATSE data. Comparing 10 A-GRBs to a ``main'' sample of about 500 bursts, we find there is a probability of only 0.03 of the samples being drawn from the same population, with the A-GRBs tending to have relatively less power on sub-second timescales. We conclude that this result is tentative evidence for variations in the properties of GRB progenitors that affect both the gamma-ray and afterglow properties of long-duration GRBs. In addition, our method of analyzing variability identifies a characteristic timescale of ~1 s, below which variability is suppressed, and finds a trend of increased short timescale variability at higher gamma-ray energies. The long-duration GRBs that we identify as having the most sub-second timescale variability, may be relatively bright examples of short-duration GRBs.
I discuss two questions about the origin of star clusters: How long does the process take? What is the mode of individual star formation? I argue that observations of Galactic star-forming regions, particularly the Orion Nebula Cluster (ONC), indicate that cluster formation often takes several Myr, which is many local dynamical timescales. Individual stars and binaries, including massive stars, appear to form from the collapse of gas cores.
We attempt to explain the properties of the Becklin-Neugebauer (BN) object as a runaway B star, as originally proposed by Plambeck et al. (1995). This is one of the best-studied bright infrared sources, located in the Orion Nebula Cluster -- an important testing ground for massive star formation theories. From radio observations of BN's proper motion, we trace its trajectory back to Trapezium star theta^1C, the most massive (45 Msun) in the cluster and a relatively tight (17 AU) visual binary with a B star secondary. This origin would be the most recent known runaway B star ejection event, occurring only \~4000 yr ago and providing a unique test of models of ejection from multiple systems of massive stars. Although highly obscured, we can constrain BN's mass (~7 Msun) from both its bolometric luminosity and the recoil of theta^1C. Interaction of a runaway B star with dense ambient gas should produce a compact wind bow shock. We suggest that X-ray emission from this shocked gas may have been seen by Chandra: the offset from the radio position is ~300 AU in the direction of BN's motion. Given this model, we constrain the ambient density, wind mass-loss rate and wind velocity. BN made closest approach to the massive protostar, source ``I'', 500 yr ago. This may have triggered enhanced accretion and thus outflow, consistent with previous interpretations of the outflow being a recent (~10^3 yr) "explosive" event.
It is well known that the energy input from massive stars dominates the thermal and mechanical heating of typical regions in the interstellar medium of galaxies. These effects are amplified tremendously in the immediate environment of young massive star clusters, which may contain thousands of O and B stars. We present models of star cluster formation that attempt to account for the interplay between feedback and self-gravity in forming clusters, with application to systems ranging from the Orion Nebula Cluster, the Galactic Center clusters like the Arches, and super star clusters.
We propose that supermassive stars may form in quasar accretion disks, and we discuss possible observational consequences. The structure and stability of very massive stars are reviewed. Because of high accretion rates, quasar disks are massive and the fringes of their optically luminous parts are prone to fragmentation. Starting from a few hundred solar masses, a dominant fragment will grow to the isolation mass, which is a significant fraction of the disk mass, more quickly than the fragment contracts onto the stellar main sequence. A gap will form in the disk and the star will migrate inward on the accretion timescale, which is comparable to the star's main sequence lifetime. By interrupting the gas supply to the inner disk, the gap may temporarily dim and redden the quasar. The final stages of stellar migration will be a strong source of low-frequency gravitational waves.
We consider the absorption and scattering of X-rays observed from the Galactic center. One objective is to characterize the intrinsic X-ray emission from the central black hole, SgrA*, in its quiescent and flaring states. We correct the fluxes observed by the Chandra and XMM telescopes for absorption and scattering, using realistic models for the dust grain size distribution, gas and dust abundances and spatial distributions. However, there are still large uncertainties. Since much of the intervening dust is close to SgrA*, the scattered halo of X-ray photons is very concentrated: its intensity can dominate the Chandra PSF inside 1", and so affects estimates of the point source flux. It also broadens the radial intensity profiles of Galactic center sources - observations of this effect can help constrain models of the spatial distribution of the dust. We estimate that the combined scattering halos from observed Galactic center sources within 10" of SgrA* contribute up to ~10% of the observed diffuse emission in this region. Dust-scattered photons suffer a time delay relative to the photons that arrive directly. For dust that is 100 pc towards us from the Galactic center, this delay is about 1000 s at angles of 1", and 100 ks at 10". We illustrate how the evolution of the scattering halo following X-ray flares from SgrA* or other sources can also help to constrain the dust's line of sight distribution. We discuss the implications of X-ray scattering halos for the intensity of diffuse emission that has been reported within a few arcseconds of SgrA*: in the most extreme, yet viable, model we consider, ~1/3 of it is due to dust scattering of an unresolved source. The remainder results from an extended source of emission.
Magnetically mediated disk outflows are a leading paradigm to explain winds and jets in a variety of astrophysical sources, but where do the fields come from? Since accretion of mean magnetic flux may be disfavored in a thin turbulent disk, and only fields generated with sufficiently large scale can escape before being shredded by turbulence, in situ field production is desirable. Nonlinear helical inverse dynamo theory can provide the desired fields for coronae and outflows. We discuss the implications for contemporary protostellar disks, where the MRI (magneto-rotational instability) can drive turbulence in the inner regions, and primordial protostellar disks, where gravitational instability drives the turbulence. We emphasize that helical dynamos are compatible with the magneto-rotational instability, and clarify the relationship between the two.
Are magnetic fields important in primordial star formation? Assuming that star formation occurs via an accretion disk that is turbulent, initially because of local gravitational instability, we calculate the disk structure for realistic accretion rates. We predict that local gravitational viscosity is able to drive accretion, without the disk fragmenting. We then estimate the rate of dynamo amplification of seed magnetic field. Turbulence in a stratified disk can be helical, with different signs of the helicity in each hemisphere. This provides a key ingredient for production of global scale magnetic fields whose sign of flux is sustained over many orbit times. The resulting fields can drive collimated protostellar outflows that reduce the star formation efficiency from the initial gas cloud, especially once the protostar has contracted to the main sequence, at ~100 Msun. We estimate that the outflows are powerful enough to eject some material from the host dark matter halo and to initiate relatively strong magnetization of the local intergalactic medium. Close to the protostar, the outflow acts to shield the disk and equatorial regions from radiative feedback, such as ionizing photons, and this may enable accretion up to relatively large stellar masses. We conclude that magnetic fields cannot be ignored from models of primordial star formation.
We present a theoretical model for primordial star formation. First we describe the structure of the initial gas cores as virialized, quasi-hydrostatic objects in accord with recent high resolution numerical studies. The accretion rate can then be related to characteristic densities and temperatures that are set by the cooling properties of molecular hydrogen. We allow for rotation of the gas core, assuming angular momentum conservation inside the sonic point of the flow. In the typical case, most mass then reaches the star via an accretion disk. The structure of the inner region of this disk is described with the standard theory of viscous disks, but with allowance for the substantial energies absorbed in ionizing and dissociating the gas. The size of the protostar and its luminosity depend upon the accretion rate, the energetics of the accreting gas, and the ability of the radiation to escape from the stellar accretion shock. We combine these models for the infall rate, inner disk structure, and protostellar evolution to predict the radiation field that is the basis for radiative feedback processes acting against infall (Paper II). For realistic initial angular momenta, the photosphere of the protostar is much smaller and hotter than in the spherical case, leading to stronger radiative feedback at earlier stages in the evolution. In particular, once the star is older than its Kelvin-Helmholtz time, contraction towards the main sequence causes a rapid increase in ionizing and far-ultraviolet luminosity at masses ~30Msun in the fiducial case. Since the cores out of which the first stars formed were much more massive than 30Msun and since feedback is dynamically unimportant at lower masses, we conclude that the first stars should have had masses >~30Msun.
I compare theoretical models of massive star formation with observations of the Orion Hot Core, which harbors one of the closest massive protostars. Although this region is complicated, many of its features (size, luminosity, accretion disk, HII region, outflow) may be understood by starting with a simple model in which the star forms from a massive gas core that is a coherent entity in approximate pressure balance with its surroundings. The dominant contribution to the pressure is from turbulent motions and magnetic fields. The collapse can be perturbed by interactions with other stars in the forming cluster, which may induce sporadic enhancements of the accretion and outflow rate.
If massive stars form by disk accretion, then bipolar outflows should be generated as in the case of low-mass star formation. High accretion rates lead to high outflow rates and make the wind density very large near the protostar. We therefore predict that massive protostars have very small, jet-like HII regions confined by their outflows; we identify these HII regions with "hypercompact" (<0.01 pc) HII regions. Their lifetime is approximately equal to the accretion timescale (~10^5 yr), much longer than the sound-crossing and dynamical timescales of the ionized region. We present an analytic description of the density distribution of the outflow, relate the overall mass loss rate to the accretion rate, and normalize to values appropriate to massive protostars. For a given ionizing luminosity, we calculate the extent of the HII region and its radio spectrum. A detailed comparison is made with observations of radio source "I" in the Orion Hot Core. The spectra and morphologies of many other sources are broadly consistent with this model. We argue that confinement of ionization allows disk and equatorial accretion to continue up to high masses, potentially overcoming a major difficulty of standard accretion models.