THIS INFORMATION IS OBSOLETE AND IS PROVIDED ONLY FOR ITS HISTORICAL VALUE

by K. D. Gibson and H. A. Scheraga, Baker Laboratory of Chemistry, Cornell University, Ithaca, New York 14853

This program minimizes the energy of a crystal structure composed of rigid molecules without imposing any constraint on the structure other than the existence of a lattice. The six-lattice parameters and the 6Z rigid-body parameters of the Z bodies in the unit cell are taken as independent variables for the purposes of energy minimization. The energy is assumed to be a sum of pairwise interatomic interactions, each of which depends only on the distance between the atoms. Potentials of the 1-6-12 (plus 1-10-12 for hydrogen bonds) are mutually exclusive. An optional term describing the effect of pressure can be added to the energy in the form proposed by Busing and Matsui.

The program makes use of an efficient modern secant- type minimization routine (Gay's routine SUMSL) and an observation by Levitt concerning the computation of energy gradients with respect to generalized coordinates to achieve rapid and efficient minimization of the energy. Possible symmetry relations between the molecules in the unit cell are computed before and after energy minimization, account being taken of any point group symmetry inherent in the molecular structure. Agreement between the initial and final symmetry relations indicates that the space group has been preserved or restored during energy minimization; otherwise, the space group has changed, and the new space group must be deduced from the final symmetry relations.

The program allows for variation of the initial lattice constants and positions of the molecules in a conformal manner by requiring that the fractional coordinates of the centroid of each molecule in the asymmetric unit be specified. When the altered lattice is generated (by applying the symmetry operations of the observed space group to the molecules in the asymmetric unit) the fractional coordinates of the centroids of the molecules in the unit cell have the same values they have in the observed unit cell. In this way, changes in the input lattice parameters lead to proportional changes in the starting positions of the molecules.

Lines of Code: 12,000 FORTRAN 77