Abstract All simulations present common trends. Metal enrichment is very patchy, with rare, unpolluted regions surviving at all redshifts, inducing the simultaneous presence of metal-free and metal-rich star formation regimes. As a result of the rapid pollution within high-density regions due to the first SN/pair instability SN, local metallicity is quickly boosted above the critical metallicity for the transition. For this reason, pop III stars dominate only during the very first stages of structure formation, with an average contribution to the total star formation rate that reaches a constant value of ~10 -3 at redshift

The primordial (population III) stars are formed out of a pristine environment, where the cooling agents are limited to primordial, H-based molecules only which are able to cool the gas down to temperatures ~ 10 2 K. Therefore the mass of primordial stars should be relatively large and their spectrum is commonly referred to as ‘top- heavy’ initial mass function. These features imply very short lifetimes (up to ~10 6 yr only) and final death mostly into black holes. The only mass range where primordial stars can explode as pair-instability supernovae (PISN) and pollute the surrounding medium is [140, 260 M solar ]. The subsequent star formation events in enriched regions will happen in completely different conditions, because metals allow further cooling and fragmentation to smaller scales.

Fate of Population III stars as a function of initial mass

Conclusions and Discussion The main influence of chemical evolution on the following generations of stars is via metal pollution. This event can completely alter the cooling properties of the gas and thus the modalities of star formation, inducing a transition from top-heavy to a standard IMF. There is assumed to be a threshold metallicity, z crit, above which which all further star-forming populations will transition from an IMF characteristic of Population III stars to that of Population II stars. The values for z crit chosen in this study include 10 -6, , and Z solar

The gas is easily enriched above z crit. For this reason, the average contribution from pristine, metal-free stars to the total cosmic star formation density is dominant only in the very early phases of structure formation, while it drops below ~10 -3 quite rapidly, after the explosion of the first PISN and their metal ejection. PISN explosions which follow the death of the metal- free or very-metal-poor stars, are responsible for enriching the surrounding medium up to a minimum level of ~10 -4 Z solar. Only minimal differences are found upon the selection of differing values for z crit, differences are only found if different IMF mass ranges for primordial stars are used.

Because of the short life and high metal yields of early, massive SN, the pop III regime contributes slightly to the global star forming rate since the early pollution events quickly raise Z above z crit. The results clearly predict that pop III star formation reaches a constant value of ~10 -3 at redshift z ~ This is a solid result, which holds independently from many poorly constrained parameters. The simulations were performed using a standard ∧ CDM cosmology, but slightly different results would not change the general picture. After ~2 x10 8 yr, molecular evolution leads to the very first bursts of star formation (pop III), but metal enrichment is extremely fast leading to pop II IMF.

Metal pollution proceeds from the densest cores of star formation outwards, because of SN ejections from high-density to lower-density environments. Rare unpolluted regions can still survive, determining the simultaneous presence of two star formation regimes, and Z crit can affect the level of residual pop III star formation. As a result, we find that the average contribution of pop III component to the total star formation rate density is a few x or by z ~ 11.

Metallicity maps at redshift z = 11 Maio et al. Figure 1

Comparison of metallicity maps at redshift z = 11, 12 and 13 for Zcrit = and 10 -5

Metal evolution as a function of redshift, Maio,et al. Figure 4

Because of the short life and high metal yields of early, massive SN, the pop III regime contributes slightly to the global star forming rate since the early pollution events quickly raise Z above z crit.

Part II – Looking ahead First Light Sources at the End of the Dark Ages: Direct Observations of Population III Stars, Proto- Galaxies, and Supernovae During the Reionization Epoch -A White Paper Submitted to the Astro2010 Decadal Survey Committee; Cooke et al. Although the initial epoch of star formation has long been a topic of theoretical interest, technology is only now beginning to allow observational insights into this epoch.

Strongest direct spectroscopic signature of these hot stars in the observed-frame optical and near-infrared is the likely intense He II (1640 A) emission, which would indicate a hard ionizing radiation field typical of the top-heavy IMF characteristic of Pop III star formation. However, this feature could also be excited by AGN activity or by accretion on to mini-black holes, which could confuse the interpretation. Model spectral energy distributions of massive, metal-free (Population III) stars from Schaerer & Pello showing the wealth of signatures of the “hard” radiation field produced by these massive and hot stars.

Pair-Production Supernovae Detection Second method involves detection of Pop III pair-instability supernovae (PISNe; e.g. Heger & Woosley 2002); these might be visible in their rest-frame ultraviolet continua with JWST. PISNe energies on the order of – ergs may allow them to be distinct and readily detectable with the GSMTs. Spectroscopy of the resultant supernovae with ground-based Giant Segmented Mirror (infrared) Telescopes (e.g., GSMT) should reveal metal abundance patterns typical of these supernovae providing unambiguous confirmation of the Pop III nature of the progenitor stars.

Observing Pop III PISNe At least a fraction of Type IIn supernovae (SNe IIn) are energetic PISNe, providing the only observable examples of this process. They have potential to yield enormous insight into the behavior of high-mass (> 140 Msolar) Pop III stars. Moreover, SNe IIn are the most luminous SN type in the rest- frame UV, rendering them easier to detect at high redshift than any other SN type. SNe IIn are defined by the presence of extremely luminous, long-lived emission lines that are dominated by Lyman α, MgII, and Hα. The combination of sensitive broad-band near-infrared imaging and spectroscopic follow up with JWST, JDEM or ground-based GSMTs will provide the means to distinguish PISNe from core-collapse SNe by their light curve rise times, overall energy output, and relative emission-line characteristics.

Emission-line strength evolution of local SNe IIn redshifted to z~6 from Cooke (2008). At z~6 and beyond, the emission lines of SNe IIn may be detectable using meter class GSMTs. Observations require a baseline of 2-10 yrs – not target of opportunity observations.

Direct Detection of Population III “Dwarf” Galaxies At “modest” redshifts (z<11) where Lyman α emission appears in the J-band, one of the best environments to discover Pop III dwarf galaxies may be the chemically unevolved surroundings of a large galaxy or proto-cluster of galaxies. The central sources are likely to have ionized a local “bubble” through which strong Lyman α emission can escape from surrounding dwarf galaxies. The near-infrared observations of these dwarfs could reveal strong Lyman α and He II emission detectable by large ground- based telescopes, and possibly a rest-frame ultraviolet continuum observable from the ground and/or with the JWST. When strong Lyman α emitters are found, both the Lyman α and He II lines can be observed with R>3000 spectroscopy using a GSMT.

Direct Detection of Pop III “Dwarf Galaxies In Population III sources, direct detection of the He II emission line, expected with equivalent widths that exceed ~10 A, is within the reach of ground-based meter telescopes. The figure on the following page illustrates simple predictions for the He II counts of various star-forming scenarios described in Barton et al. (2004). The “optimistic” scenario corresponds to pure Pop III, metal-free, solar-mass stars. Both the optimistic and “plausible” scenarios can be detected with next-generation large ground-based telescopes.

Simple model estimates of the He II (1640 A) “luminosity function at z ~ 8. Following Barton et al. (2004), we plot the expected source counts in He II per square arcminute per redshift increment near z ~8 for fully Population III (“optimistic”), low-metallicity (“plausible), and heavy Population II (“heavy Salpeter”) scenarios. We also indicate show the approximate sensitivity range of the GSMTs. The presence of strong, detectable HeII emission is a clear indication of a drastic change in the stellar initial mass function and metal content of high-redshift galaxies.

We can use the results from JWST and/or ground-based infrared surveys to search for intense Lyman α sources. When strong Lyman α emitters are found, both the Lyman α and the He II lines can be observed with R > 3000 spectroscopy using a GSMT. Observations can focus on the region in which He II is expected (e.g., at 1.44 microns or H- band, for the z ~ 7.7 window). If a He II (1640 A) feature is discovered that is strong relative to the UV continuum limits from JWST and the Lyman α feature (e.g., rest frame EW (He II) > A), the strength of the feature provides direct evidence that the source is not a “standard” Pop II star formation region similar to what is observed in the local universe. (as opposed to Wolf-Rayet stars with EW of a few Angstroms)