Texas Stellar Archaeology

Probing the formation of the first low-mass stars!

Emergence from the Dark Ages

We live in a universe that is full of bright objects. According to modern cosmology, however, the universe was featureless and dark for a long time of its early history. The first stars did not appear until about 100 to 250 milion years after the Big Bang. They formed in a what is called a minihalo of 100,000 to one million times more massive than the sun and measuring 30 to 100 light years across. Those minihaloes evolved from density fluctuations in the early universe. Recent cosmological simulations conclude that the first stars in the universe were typically 100 times more massive and luminous than the sun.

Cosmic Timeline and evolution of stars after Big Bang

According to the modern cosmology, all space, time, and energy began with the Big Bang which occured about 13.7 billion years ago. The first stars in the Universe are predicted to have been extremely massive, about 100 times heavier than the sun. Gravitational collapse, accompanied by cooling of the primordial gas via molecular hydrogen, yields a minimum fragmentation scale of a few hundred solar masses. From the numerical simulations we know that once a gas clump acquires this mass it undergoes a slow, quasi-hydrostatic contraction but without further significant fragmentation. Lower-mass stars such as the sun cannot form in the priordial environment.

The time span from Big Bang to the 1st generation of stars is currently not accessible by observations. Hence, astronomers heavily rely on theoretical studies. By observing low metallicity stars of the 2nd generation, however, we can check whether our observations match any theoretical predictions about the formation of those early low-mass stars. We call those early generation stars "stellar fossils" because they provide us with information of the earliest times in the Universe. Because of their low-mass, these stars have burnt their fuel only very slowly are are thus still observable today in the our own Milky Way Galaxy.

"Probing the formation of the first low-mass stars with stellar archaeology"

by Frebel, Johnson & Bromm (2007, MNRAS 380, L40) addresses why it is important to probe the formation of first low-mass stars.

Abstract:

We investigate the conditions under which the first low-mass stars formed in the Universe by confronting theoretical predictions governing the transition from massive Population III to low-mass Population II stars with recent observational C and/or O abundance data of metal-poor Galactic stars. We introduce a new "observer-friendly" function, the transition discriminant D_trans which provides empirical constraints as well as a powerful comparison between the currently available data of metal-poor halo stars and theoretical predictions of the formation of the first low-mass stars (less than 1 solar mass) . Specifically, we compare the empirical stellar results with the theory that fine-structure lines of C and O dominate the transition from Population III to Population II in the early Universe. We find the currently available data for halo stars as well as for dwarf spheroidal (dSph) galaxies and globular clusters to be consistent with this theory. An explanation for the observed lack of metal-poor stars in dSph galaxies and globular clusters is also suggested. Finally, we predict that any star to be found with [Fe/H] <-4 should have enhanced C and/or O abundances. The high C and O abundances of the two most iron-poor stars are in line with our prediction.

Two cooling models that could be responsible for the transition of the massive first stars (Population III) to low-mass stars (Population II)

Currently, there are two competing models for the Pop III to Pop II transition discussed in the literature: (i) atomic fine-structure line cooling (Bromm & Loeb 2003; Santoro & Shull 2006); and (ii) dust-induced fragmentation (e.g. Schneider et al. 2006). Within the fine-structure model, C II and O I have been suggested as main coolants (Bromm & Loeb 2003), such that low-mass star formation can occur in gas that is enriched beyond the critical abundances of [C/H] _crit ~ −3.5 ± 0.1 dex and [O/H] _crit ~ −3 ± 0.2 dex. The dust-cooling model, on the other hand, predicts critical abundances that are typically smaller by a factor of 10–100. This order-of-magnitude difference is much larger than the uncertainties within the fine-structure line model itself, which are still substantial, regarding the precise value of the critical abundances. Also, elements other than C and O (such as Si and Fe) might contribute to the cooling (e.g. Santoro & Shull 2006), but are unlikely to exceed a factor of a few in the overall cooling rate. In order to find the oldest, most "primordial" low-mass stars, Bromm & Loeb (2003) suggested searching for the most C- and O-poor objects. Such abundances should reflect the level of primordiality of a star independently of the abundances of heavier elements (e.g. Fe). They may thus provide an important and feasible observational tool to constrain the formation mechanisms of the first low-mass stars in the high-redshift Universe (e.g. Salvadori et al. 2007) while also offering an empirical discrimination between the two cooling models. Such constraints, in turn, should also provide clues to the relevance of the first generations of stars to cosmic reionization (e.g. Tumlinson, Venkatesan & Shull 2004). Detailed knowledge about the first epoch of star formation will also allow improved predictions for future high-redshift observations (e.g. Barkana & Loeb 2007). Together with the currently available data, the imminent arrival of large samples of metal-poor stars such as from SEGUE, LAMOST, and the Skymapper survey, provides an ideal test-bed to empirically constrain and possibly differentiate between different cooling theories.

The Transition Discriminant D_trans

We introduced a new function, the "transtion discriminant". It is defined as follows:

Low-mass star formation requires D_trans > D_trans,crit= −3.5. This function reproduces the critical values in Bromm & Loeb (2003) for the cases where only C or O are contributing to the cooling. It approximates the case where both elements are present by assuming that the individual cooling rates are simply added. The D_trans function now has the important advantage that cosmologically relevant measurements can be obtained from stellar C and/or O abundances. In particular also from the common case where only one of the two abundances is measured in a star can be accomodated by this new function.

In the figure shows D_trans as a function of metallicity [Fe/H], a measure for the look-back time into the early Universe. In the top panel, we test D_trans,crit with Milky Way halo stars. We note that most of the data are lower limits because of missing O abundances. The C-based lower limits, however, are much more robust than those derived from O because D_trans, by definition, is dominated by the C abundance. Hence, we still derive cosmologically relevant measures just from the C abundance. All stars lie above the "forbidden zone" (as marked in the figure) or within a typical observational error (0.2 to 0.3 dex) of the critical limit of D_trans,crit. There appears to be a lower envelope of the overall distribution of the data. Keeping in mind that many of the values are lower limits, this envelope depicts the solar C and O abundances (solid line).

In the lower panel of figure we show stars from globular cluster and dwarf spheroidal (dSph) galaxies collected from the literature. Most of the stars only have an O abundance available. With one exception, none of them has a metallicity less than [Fe/H] ~−2.5. In the absence of low-metallicity stars, these dSph systems must have had available sufficient amounts of at least C and O at an early time to avoid the formation of very metal-poor stars.

New project:

We have finished a new project to derive observational constraints on the initial mass function of the first stars (Pop III). We also determine the accretion rate of metal-poor stars to show that stellar archaeology works that that the oldest, most metal-poor stars really do tell us about the conditions of the early Universe. Stay tuned!

Related movies

Hydrogen density interaction with high-energy, ionizing photons during the formation and decay of eight stars.
(54 seconds)

Molecular density interaction with high-energy, ionizing photons during the formation and decay of eight stars.
(21 seconds)