Nucleosynthesis in the first massive stars

The nucleosynthesis in the first massive stars may be constrained by observing the surface composition of long-lived very iron-poor stars born around 10 billion years ago from material enriched by their ejecta. Many interesting clues on physical processes having occurred in the first stars can be obtained based on nuclear aspects. First, in these first massive stars, mixing must have occurred between the H-burning and the He-burning zone during their nuclear lifetimes; Second, only the outer layers of these massive stars have enriched the material from which the very iron-poor stars, observed today in the halo of the MilkyWay, have formed. These two basic requirements can be obtained by rotating stellar models at very low metallicity. In the present paper, we discuss the arguments supporting this view and illustrate the sensitivity of the results concerning the [Mg/Al] ratio on the rate of the reaction ²³Na(p,γ)²⁴Mg. (paper)