Development of a Phenotype Algorithm for Particle Geometry Optimization

We implemented a search algorithm capable of finding the energetically most favorable and thus stable spatial configurations of particles interacting pairwise via potential functions, considering two different types of phenomena, namely cluster and lattice formation. This problem was successfully solved using concepts of genetic algorithms, in particular so-called phenotype algorithms introduced in the late 1990s. With the help of our algorithm, we gained results on well-known benchmark systems (including Lennard-Jones Clusters) as well as on a newly proposed two-dimensional system.