A tabu-search based strategy for modeling molecular aggregates and binary reactions.

Modern day computational chemistry is moving towards studying complex chemical systems, and global optimization algorithms are becoming increasingly important in successfully carrying out such studies. Herein, we report on a simple tabu-based strategy that generates sufficiently dissimilar trial structures for subsequent optimization, in which the minimum degree of similarity between initial trial structures is a user-definable parameter. This versatile tabu-based strategy was implemented for modeling aggregation as well as chemical reactions. The method was validated on molecular aggregates of water, CO 2 , and acetic acid, and was successful in finding the global minima and several of the important low-lying minima on the potential energy surface. This strategy was further extended to the efficient generation of trial geometries as applied to chemical reactions, and was validated by predicting the products of binary reactions using a set of 10 Diels-Alder reactions in the gas-phase. After the successful validation, our method was applied to the prediction of various products of Diels-Alder reactions on graphene models. [ABSTRACT FROM AUTHOR]