Computational Study of the Structure and Thermodynamic Properties of Ammonium Chloride Clusters Using a Parallel J-Walking Approach

Abstract

The thermodynamic and structural properties of (NH$_4$Cl)$_n$ clusters, n=3-10 are studied. Using the method of simulated annealing, the geometries of several isomers for each cluster size are examined. Jump-walking Monte Carlo simulations are then used to compute the constant-volume heat capacity for each cluster size over a wide temperature range. To carry out these simulations a new parallel algorithm is developed using the Parallel Virtual Machine (PVM) software package. Features of the cluster potential energy surfaces, such as energy differences among isomers and rotational barriers of the ammonium ions, are found to play important roles in determining the shape of the heat capacity curves.