Prediction of energy structure of three dimensional born clusters based on dynamic lattice optimization algorithm

Objective Cluster science is a very important research field in computational chemistry, and the prediction of the lowest energy structure of B40 and B45 has always been a difficult problem. However, the traditional technology can't predict the structure well, so it is necessary to use artificial intelligence technology for intelligent prediction. Different from the traditional iterative Schrodinger equation, this paper proposes an improved Lennard Jones model and uses computer algorithm to predict the global optimal structure. Methods Based on the L-J model, this paper introduces the improved L-J model, and uses the ejection method and dynamic lattice optimization algorithm to generate the initial configuration randomly. Results Compared with the traditional model, the improved L-J model was better, and the goodness of fit reached 97.3%. In addition, the predicted global optimal structure of B40 is hollow body, and the minimum energy is - 84475.5; The global optimal structure of B45 is sphere like, and the minimum energy is −136395. Conclusion Compared with the traditional L-J model, the model generated by artificial intelligence algorithm can better predict the lowest energy configuration of boron clusters, which is universal.