Positron states and annihilation rates were calculated for 9 different II-VI compound semiconductors. Positron annihilation from delocalized states in the perfect lattice, as well as from localized states at vacancies and di-vacancies, was considered. The calculations were based upon the local-density approximation for electron-positron correlation effects. The calculations were performed by using non self-consistent electron densities and electrostatic potentials which were obtained by atomic superposition, and by solving for the 3-dimensional positron wave-functions by using a relaxation method. In the case of perfect lattices, self-consistent electron densities and positron states (linear muffin-tin orbital method within atomic spheres approximation) were also calculated. The results showed that positron annihilation with the outer d-electrons of the group-II metal atoms played an important role. The positron lifetimes which were calculated for perfect lattices were a few percent shorter than the experimental ones. This indicated a local-density over-estimation of the d-electron enhancement around the positron. These bulk lifetimes could be corrected efficiently by using a semi-empirical model. In the case of positrons which were trapped by defects, the present theoretical description was less satisfactory, because atomic relaxations which were due neither to rearrangements in the electronic structure nor to the localized positrons were taken into account. However, the calculated annihilation probabilities with core and valence electrons were expected to serve as an important basis for methods in which defects could be identified by using positron angular correlation or Doppler broadening measurements.

F.Plazaola, A.P.Seitsonen, M.J.Puska: Journal of Physics - Condensed Matter, 1994, 6[42], 8809-27