An ab initio study of the doping of alkali halide crystals (AX: A=Li, Na, K, Rb; X = F, Cl, Br, I) by ns^2
anions (Ag^- and Cu^-) is presented. Large active clusters with 179 ions embedded in the surrounding crystalline
lattice are considered in order to describe properly the lattice relaxation induced by the introduction of
substitutional impurities. In all the cases considered, the lattice distortions imply the concerted movement of
several shells of neighbors. The shell displacements are smaller for the smaller anion Cu^-, as expected. The
study of the family of rock-salt alkali halides (excepting CsF) allows us to extract trends that might be useful
at a predictive level in the study of other impurity systems. Those trends are presented and discussed in terms
of simple geometric arguments.
