Pre-supernova mixing in CEMP-no source stars

Context. CEMP-no stars are long-lived low-mass stars with a very low iron content, overabundances of carbon and no or minor signs for the presence of s- or r-elements. Although their origin is still a matter of debate, they are often considered as being made of a material ejected by a previous stellar generation (source stars). Aims. We place constraints on the source stars from the observed abundance data of CEMP-no stars. Methods. We computed source star models of 20, 32, and 60 M-circle dot at Z = 10(-5) with and without fast rotation. For each model we also computed a case with a late mixing event occurring between the hydrogen and helium-burning shell similar to 200 yr before the end of the evolution. This creates a partially CNO-processed zone in the source star. We use the C-12/C-13 and C/N ratios observed on CEMP-no stars to put constraints on the possible source stars (mass, late mixing or not). Then, we inspect more closely the abundance data of six CEMP-no stars and select their preferred source star(s). Results. Four out of the six CEMP-no stars studied cannot be explained without the late mixing process in the source star. Two of them show nucleosynthetic signatures of a progressive mixing (due e.g. to rotation) in the source star. We also show that a 20 M-circle dot source star is preferred compared to one of 60 M-circle dot and that likely only the outer layers of the source stars were expelled to reproduce the observed C-12/C-13. Conclusions. The results suggest that (1) a late mixing process could operate in some source stars; (2) a progressive mixing, possibly achieved by fast rotation, is at work in several source stars; (3) similar to 20 M-circle dot source stars are preferred compared to similar to 60 M-circle dot ones; and (4) the source star might have preferentially experienced a low energetic supernova with large fallback.