New Algorithms for Locating Global Minima of Molecular Clusters: A Progress Report and Test Applications to Water Clusters (H[sub 2]O)[sub n], n≤34
Sergey Kazachenko, Ajit J. Thakkar, George B. Maroulis, Theodore E. Simos · AIP conference proceedings · 2009
Finding the global minimum of a potential energy surface of a molecular cluster is a difficult problem because the number of minima increases exponentially with respect to the number of atoms in the system. Minima hopping is a recently developed global optimization method which offers several advantages over other popular methods. As a first step this method was implemented in TINKER. To improve performance, rotational and translational perturbations of the monomers were introduced as additional steps in the algorithm. The hydrogen bond topology plays an important role in water clusters, but minima hopping does not always find the best topology. Hence, a hydrogen bond topology optimization algorithm for water clusters was developed. We find that topology optimization can lower the energy of a cluster by significant amounts. Our algorithms were applied to pure water clusters ranging from (H2O)6 to (H2O)34 using the TIP4P intermolecular potential. Our results agree with previous work for the clusters up to (H2O)24 confirming the validity of our algorithms. Moreover, we found a better candidate for the TIP4P global minimum of (H2O)25 and report the first candidates for the TIP4P global minima of (H2O)26 through (H2O)34.