2009 Bromm et al., The formation of the first stars and galaxies (1119309), страница 3
Текст из файла (страница 3)
These simulations have shown thatearly enrichment is very inhomogeneous, as the low-density voids areenriched before any metals can reach into the denser filaments andvirialized haloes71.Assembly of the first galaxiesFigure 4 | Radiative feedback around the first stars. Ionized bubbles areshown in blue, and regions of high molecule abundance in green. The largeresidual free electron fraction inside the relic H II regions, left behind afterthe central star has died, rapidly catalyses the reformation of molecules.
Theabundance of HD molecules allows the primordial gas to cool to thetemperature of the CMB, possibly leading to the formation of populationIII.2 stars after these regions have re-collapsed so that gas densities aresufficiently high again for gravitational instability to occur77. The latterprocess takes of the order of the local Hubble time, thus imposing a,100 Myr delay in star formation.
The relatively high molecule abundancein relic H II regions, along with their increasing volume-filling fraction, leadsto a large optical depth to Lyman–Werner photons over physical distances ofthe order of several kiloparsecs (ref. 47). The development of a high opticaldepth to Lyman–Werner photons over such short length-scales, combinedwith a rapidly increasing volume filling fraction of relic H II regions, suggeststhat the optical depth to Lyman–Werner photons over cosmological scalesmay be very high, acting to suppress the build-up of a backgroundLyman–Werner radiation field, and mitigating negative feedback on starformation75.
Note the strongly clustered nature of early star formation.Visualization courtesy of the Texas Advanced Computing Center (based ondata from ref. 47).elements. Population III stars in the range 140–260 M8 explode aspair-instability supernovae, which disrupt the entire progenitor, withexplosion energies ranging from 1051 erg to 1053 erg, and nucleosynthetic yields, defined as the heavy-element mass fraction, up to 0.5.
Suchsupernovae exhibit an odd-even effect in the nuclei produced that ismuch greater than observed in any star to date, and as a result theycannot make a significant contribution to the metals observed in verylow-metallicity stars today57. On the other hand, the pair-instabilitysupernova signature may exist in a tiny fraction of the stars with intermediate metallicity (,0.01Z8 , where Z8 indicates solar metallicity),because the enrichment from even a single pair-instability supernovaalready endows the surrounding material with heavy elements to levelsthat are above the regime typically probed by surveys of metal-poorstars58.The first stars may have been born rapidly rotating, however, androtation can entirely modify these results59.
For sufficiently highrotation rates, rotationally induced mixing is able to render the coreschemically homogeneous; mixing of heavy elements to the surface inthe late stages of evolution can lead to substantial mass loss. If thecores maintain a sufficiently high rotation at the time of the supernova, it is possible to produce a long c-ray burst or a jet-inducedenergetic supernova/hypernova60,61, with significant effects on theabundances of the ejected metals62.
Large uncertainties remain inthe evolutionary calculations owing to the effects of dynamo-generated magnetic fields.The strong mechanical and chemical feedback effects exerted byexplosions of population III stars have been investigated with a number of detailed calculations63–69. The key question is how the initiallymetal-free Universe was enriched with the first heavy chemical elements70. Recently, it has become feasible to address this process withThe characteristic mass of the first star formation sites has beendetermined to be ,106M8 (refs 14, 72), whereas the critical massfor hosting the formation of the first galaxies is still not known withany certainty. A promising theoretical Ansatz is to explore atomiccooling haloes—with ,108M8 and virial temperatures greater than,104 K so that atomic line cooling is efficient—as their formationsites73,74.
The simulations, starting from cosmological initial conditions, are just now approaching the resolution and physical realismrequired to investigate whether atomic cooling haloes fulfil the criteria for a first galaxy as defined above. Quite generically, in suchmodels, the first generation of stars forms before galaxies do, andfeedback effects from the first stars are expected to play a key role indetermining the initial conditions for the formation of the firstgalaxies. Although substantial uncertainties in the overall formationefficiency of the first stars still remain, it is possible, and perhapsprobable, that at least one primordial star had formed in the regionthat was destined to eventually become a first galaxy75. If the earlygeneration stars were massive, >10M8 , the feedback effects describedin the previous section would shape the conditions for subsequent starformation in the region.The gas expelled by the H II regions and supernovae of the first starswould have been too hot and diffuse to allow further star formationuntil it had time to cool, as well as to reach high densities again in thecourse of being reincorporated in a growing dark-matter halo.
Bothcooling and re-collapse occur rather slowly, thus rendering starformation intermittent in the early formation phase of the firstgalaxies. Analytic models76 and detailed numerical simulations47,77both show that the gas re-incorporation time is as long as 108 years,roughly corresponding to the dynamical time for a first-galaxy haloto be assembled.Chemical enrichment by the first supernovae is among the mostimportant processes in the formation of the first galaxies. Efficientcooling by metal lines and dust thermal emission regulate the temperature of already metal-enriched population II (see Box 1) star-formingregions in the first galaxies.
The concept of a ‘critical metallicity’ hasbeen introduced to characterize the transition of the star-formationmode from predominantly high-mass, population III or population II,to low-mass population II stars78. However, this critical gas metallicityis still poorly determined. It is not even clear if there exists such a sharptransition. Some studies show that even a slight quantity of metals in agas may be enough to change the gas thermal evolution significantly79,whereas others argue that the cooling efficiency at low densities80 iscrucial and is significantly enhanced only above 10{4 Z8 . As theenrichment from even a single pair-instability supernova by a verymassive population III star probably leads to metallicities ofZw10{2 Z8 (ref.
63), well in excess of any predicted value for thecritical metallicity, these arguments might be somewhat academic.The characteristic mass of pre-stellar gas clumps is probably determined by a number of physical processes (for example, turbulenceand, possibly, dynamo-amplified primordial magnetic fields) otherthan radiative cooling. The overall effect of gas metallicity on starformation may well be limited81.Recent cosmological simulations have demonstrated that starformation inside the first galaxies was strongly influenced by gravitationally driven supersonic turbulence that was generated during thevirialization process64,73,74.
This is in marked contrast to the ratherquiescent, quasi-hydrostatic situation in minihaloes (see Fig. 5). Itthus appears possible that the first galaxies harboured the first stellarclusters, if present-day star formation offers any guide here, where it52©2009 Macmillan Publishers Limited. All rights reservedREVIEWSNATUREjVol 459j7 May 2009is widely believed that gravo-turbulent fragmentation is responsiblefor shaping the initial mass function24,82. It is an open question as towhether the first galaxies could have harboured the first globularclusters, which are the oldest star clusters known.Future empirical probesStudying the formation of the first stars and galaxies will be at thefrontier of astronomy and cosmology in the next decade.Astronomers will muster a comprehensive arsenal of observationalprobes. The most prominent among these concern the optical depthto Thomson scattering of cosmic microwave background photons offfree electrons83–85, the near-infrared background86, high-redshiftc-ray bursts87–89, the possibility of scrutinizing the nature of the firststars by metals found in the oldest Galactic halo stars, dubbed ‘stellararchaeology’90,91, and various facilities now being deployed to mapreionization using the redshifted 21 cm line of neutral hydrogen92–94.The James Webb Space Telescope (JWST) will perform a number ofobservations designed to test key assumptions of our current theoryof the first stars and galaxies95.
How could the existence of massivepopulation III stars be unambiguously inferred? The most clear-cutdiagnostic is the ratio of recombination lines emitted from the H IIregions around single population III stars, or clusters thereof, to bemeasured with ultra-deep near-infrared and mid-infrared spectroscopy. Owing to the high effective temperature of the population IIIstellar continuum, ,105 K, strong He II line emission at a rest-framewavelength of 1,640 Å is predicted, with a ratio compared to Lymana that is one to two orders of magnitude larger than for normalstars41.
A second crucial observational campaign aims at a censusof very high-z supernovae96 through deep broadband near-infraredimaging. One key objective is to search for possible pair-instabilitysupernova events, which would clearly stand out owing to theirextreme intrinsic brightness, as well as their very long durations—afew years in the observer frame97. The goal of making useful predictions for the high-redshift frontier is now clearly moving withinreach, and the pace of progress is likely to be rapid.1.2.3.4.5.6.7.8.9.10.11.12.13.14.15.16.17.18.19.20.21.22.23.24.25.26.log Ma–1012Size: 40 kpc (comoving)x–y planez = 10.62tH = 429.4 MyrFigure 5 | Turbulence inside the first galaxies. Shown is the Mach number(Ma) in a slice through the central 40 kpc (comoving) of the galaxy.
Thedashed line denotes the virial radius of ,1 kpc. The Mach numberapproaches unity at the virial shock, where the accreted gas is heated to thevirial temperature. Inflows of cold gas along filaments are supersonic by afactor of ,10, resulting in strong turbulent flows in the galactic centre. Theage of the Universe at redshift z < 10 is given by tH. Reproduced bypermission of Wiley-Blackwell (from ref. 74).27.28.29.30.31.32.Barkana, R.
& Loeb, A. In the beginning: The first sources of light and thereionization of the Universe. Phys. Rep. 349, 125–238 (2001).Miralda-Escudé, J. The dark age of the Universe. Science 300, 1904–1909 (2003).Loeb, A., Ferrara, A. & Ellis, R. S. First Light in the Universe (Springer, 2008).Bromm, V. & Larson, R. B. The first stars. Annu.
![](https://s.studizba.com/z.php?f=/templates/si/i/noavatar.png&w=100&h=100&t=1"){kind=avatar}
