Abstract
It is shown that the structural phase transformations and relative stabilities among cubic phases of sp-bonded nonmetals can be successfully predicted within a simple, universal tight-binding model by a total-energy-minimization procedure. The model elucidates the physical mechanisms determining the chemical trends and predicts semiquantitatively the stable crystal structures, bond lengths, bulk moduli, and transition pressures of structural phase transformations. The theory explains the puzzling strong cation and weak anion dependence of the observed structural transition pressures.