Modified clustering algorithm for molecular simulation

FC de Oliveira and S Khani and JM Maia and FW Tavares, MOLECULAR SIMULATION, 46, 1453-1466 (2020).

DOI: 10.1080/08927022.2020.1839661

Herein, an extended Hoshen-Kopelman algorithm to non-lattice environments has been used to create a algorithm suitable for cluster detecting/labelling in molecular simulations. This algorithm is useful to study many systems of particles which undergo aggregating phenomena, such as clustering or gelation. Particle positions are input in this code, which verifies connections between molecules based on a cutoff distance. We show two examples to better illustrate the proposed algorithm: the aggregation of Lennard-Jones particles under Brownian motion at gelation condition and aggregation of asphaltene molecules in heptane. Results showed that the number of clusters diminishes because each molecule/particle initially corresponds to one cluster, which aggregates over time. Consequently, the number of links (the mean size of the cluster) increases until reaching a constant value, which means that the system achieved equilibrium. This new algorithm is compared to the previously existing one and proved to be faster, while showing to be very useful in counting the number of clusters.

Return to Publications page