Molecular geometry optimization with a genetic algorithm
D.M. Deaven and K.M. Ho
Physical Review Letters 75, p288 (1995).
Abstract: We present a method for reliably determining the lowest energy structure of an atomic cluster in an arbitrary model potential. The method is based on a genetic algorithm, which operates on a population of candidate structures to produce new candidates with lower energies. Our method dramatically outperforms simulated annealing, which we demonstrate by applying the genetic algorithm to a tight-binding model potential for carbon. With this potential, the algorithm efficiently finds fullerene cluster structures up to C_60 starting from random atomic coordinates. (PDF)
Structural optimization of Lennard-Jones clusters by a genetic
D.M. Deaven, N. Tit, J.R. Morris and K.M. Ho
Chemical Physics Letters 256, p195 (1996).
Abstract: We use a newly-developed genetic algorithm to determine the lowest energy atomic configurations of 2 -- 100 atoms in the Lennard-Jones potential. Our method, which contains no bias to specific symmetries, yields structures which are identical to or are lower in energy than all previously published structures. (PDF)
Genetic Algorithm Energy Minimization for Point Charges on a
J.R. Morris, D.M. Deaven and K.M. Ho
Physical Review B XXX, pXXX (1996).
Abstract: We demonstrate that a new approach for optimizing atomic structures is very effective for attacking the Thomson problem of finding the lowest energy configuration of N point charges on a unit sphere. Our approach uses a genetic algorithm, combined with a `cut and paste' scheme of mating, that effciently explores the different low energy structures. Not only have we reproduced the known results for 10<N<132, this approach has allowed us to extend the calculation for all N < 200. This has allowed us to identify series of `magic' numbers, where the lowest energy structures are particularly stable. Most of these structures are icosahedral, but we also find new low-energy structures that deviate from icosahedral symmetry. (PDF)