Detecting Primordial Stars

We study the detectability of primordial metal-free stars, both through direct searches for their emission, as well as searches for the resulting supernovae. We show that enrichment is a local process that takes place over an extended redshift range. While the duration of the transition from a metal-free to an enriched universe depends on several unknown factors, in all models late-forming metal-free stars are found in 107.5 M⊙ − 10 8.0 M⊙ objects, which are just large enough to cool within a Hubble time, but small enough to not be clustered near areas of previous star formation. We discuss the observational properties of these objects, some of which may have already been detected in ongoing surveys of high-redshift Lyman-α emitters. If metal-free stars have masses 140M⊙ < ∼ M⋆ < ∼ 260 M⊙, they are expected to end their lives as pair-production supernovae (PPSNe), in which an electron-positron pair-production instability triggers explosive nuclear burning. Using the implicit hydrodynamics code KEPLER, we calculate a set of PPSNe light curves that allows us to assess observational strategies for finding these objects. The peak luminosities of typical PPSNe are only slightly greater than those of Type Ia supernovae, but they remain bright much longer (∼ 1 year) and have hydrogen lines. Ongoing supernova searches may soon be able to place stringent limits on the fraction of very massive stars out to z ≈ 2. The planned Joint Dark Energy Mission satellite will be able to extend these limits out to z ≈ 6.