Using a genetic algorithm to study properties of minimum energy states and geometrical frustration in artificial “spin ice” systems

Abstract This article reports the results of a study on the base state of artificially frustrated “spin ice” systems. We have studied the states of minimum energy reported by experimental studies on nanoscale ferromagnetic islands and the protocols employed to reach those states. The main technique employed in this study is a genetic algorithm that has been contrasted with two Montecarlo methods. Nanoscale islands are modeled through dipolar moments placed on a plane, rectangular array. Studies include the correlation between nanoscale islands, statistics on vertex types formed in the array for the minimum energy state and intermediate states. The results suggest a failure in the protocols adopted to minimize energy in these systems. A study on the efficiency between the devised genetic algorithm and the Montecarlo methods used in the research is also included.