Investigation of the Stability Mechanisms of Eight-Atom Binary Metal Clusters Using DFT Calculations and k‑means Clustering Algorithm

• Published in: Journal of chemical information and modeling Vol. 61; no. 7; pp. 3411 - 3420
• Main Authors: Orlando Morais, Felipe, Andriani, Karla F, Da Silva, Juarez L. F
• Format: Journal Article
• Language: English
• Published: Washington American Chemical Society 26.07.2021
• Subjects: Algorithms; Aluminum; Cluster analysis; Clustering; Cobalt; Computational Chemistry; Copper; Correlation analysis; Density functional theory; Intermetallic compounds; Iron; Magnetic moments; Mathematical analysis; Metal clusters; Molecular orbitals; Nickel aluminides; Nickel base alloys; Nickel compounds; Stability analysis; Vector quantization
• ISSN: 1549-9596; 1549-960X; 1549-960X
• DOI: 10.1021/acs.jcim.1c00253

Here, we report density functional theory calculations combined with the k-means clustering algorithm and the Spearman rank correlation analysis to investigate the stability mechanisms of eight-atom binary metal AB clusters, where A and B are Fe, Co, Ni, Cu, Ga, Al, and Zn (7 unary and 21 binary clusters). Based on the excess energy analysis, the six most stable binary clusters are NiAl, NiGa, CoAl, FeNi, NiZn, and FeAl, and except for FeNi, their highest energetic stabilities can be explained by the hybridization of the d- and sp-states, which is maximized at the 50% composition, i.e., A4B4. Based on the Spearman correlation analysis, the energetic stability of the binary clusters increases with an increase in the highest occupied molecule orbital–lowest unoccupied molecular orbital (HOMO–LUMO) energy separation, which can be considered as a global descriptor. Furthermore, reducing the total magnetic moment values increases the stability for binary clusters without the Fe, Co, and Ni species, while the binary FeB, CoB, and NiB clusters increase their energetic stability with a decrease in the cluster radius, respectively, i.e., an energetic preference for compact structures.