Abstract

Herein the performance of a modification within the hybrid algorithm implemented in the AUTOMATON program is introduced and evaluated. For the creation of the initial population, AUTOMATON combines a probabilistic automata procedure with a genetic algorithm used to evolve this population. The proposed modification is addressed to efficiently identify the minimum energy structures of systems composed of more than one type of atom and with a low computational cost. The effectiveness of this approach is evaluated in the determination of the minimum energy structures of Be6B11-. The modification, aimed to explore the potential energy surface, consisted of filling the cells first with Be atoms in the process of creating the initial population. This order obeys the structural pattern established in the Be-B clusters reported to date. The results show that this variation not only identifies a more significant number of viable isomers but also to find a better putative global minimum than those previously reported in the literature. Therefore, it is recommended to be used as a complement to the standard searching process.