We highlight the role of the light elements (Li, Be, B) in the evolution of massive single and binary stars, which is largely restricted to a diagnostic value, and foremost so for the element boron. However, we show that the boron surface abundance i
n massive early type stars contains key information about their foregoing evolution which is not obtainable otherwise. In particular, it allows to constrain internal mixing processes and potential previous mass transfer event for binary stars (even if the companion has disappeared). It may also help solving the mystery of the slowly rotating nitrogen-rich massive main sequence stars.
We review the role of rotation in massive close binary systems. Rotation has been advocated as an essential ingredient in massive single star models. However, rotation clearly is most important in massive binaries where one star accretes matter from
a close companion, as the resulting spin-up drives the accretor towards critical rotation. Here, we explore our understanding of this process, and its observable consequences. When accounting for these consequences, the question remains whether rotational effects in massive single stars are still needed to explain the observations.
Pair creation supernovae (PCSN) are thought to be produced from very massive low metallicity stars. The spectacularly bright SN 2006gy does show signatures expected from PCSNe. Here, we investigate the metallicity threshold below which PCSN can form
and estimate their occurrence rate. We perform stellar evolution calculations for stars of 150$mso$ and 250$mso$ of low metallicity (Z$_{odot}$/5 and Z$_{odot}$/20), and analyze their mass loss rates. We find that the bifurcation between quasi-chemically homogeneous evolution for fast rotation and conventional evolution for slower rotation, which has been found earlier for massive low metallicity stars, persists in the mass range considered here. Consequently, there are two separate PCSN progenitor types: (I) Fast rotators produce PCSNe from very massive Wolf-Rayet stars, and (II) Slower rotators that generate PCSNe in hydrogen-rich massive yellow hypergiants. We find that hydrogen-rich PCSNe could occur at metallicities as high as Z$_{odot}$/3, which -- assuming standard IMFs are still valid to estimate their birth rates -- results in a rate of about one PCSN per 1000 supernovae in the local universe, and one PCSN per 100 supernovae at a redshift of $z=5$. PCSNe from WC-type Wolf-Rayet stars are restricted to much lower metallicity.
