Geometrical and electronic structures, stability, and aromaticity of copper-doped boron clusters B_nCu^0/- (n = 1-13)

Besides small pure boron clusters which have been well studied, doped boron clusters, in particular with transition metal dopants, are also increasingly investigated due to their unique geometrical structures accompanied by novel physical and chemical properties. However, studies on copper-doped boron clusters are still scarce despite copper being a transition metal with important properties. In this context, the geometry, stability, and electronic properties of copper-doped boron clusters B_nCu^0/- (n = 1-13) were systematically studied. The obtained results show that the geometries of neutral and anionic clusters are very similar only at sizes n = 2, 5, 6, 8, and 10. In the ground state, both neutral and anionic clusters tend to be in the low-spin state, except for B₂Cu^- and B₆Cu^-. The relative stability of these structures was evaluated using various energetic parameters. Moreover, the aromaticity of the B₈Cu^- anion along with its enhanced stability are also discussed and rationalized in detail.

In this work, the geometry optimization and the following molecular orbital analyses of B_nCu^0/- (n = 1-13) structures are performed using density functional theory (DFT), specifically the hybrid functional TPSSh combined with the 6-311+ G(d) basis set for boron atoms and the aug-cc-pVTZ-PP basis set for a copper atom. To obtain more reliable results, the single point energies of some lowest lying isomers whose relative energy values are close corresponding to geometries optimized at the TPSSh/6-311+ G(d) (B) / aug-cc-pVTZ-PP (Cu) level were determined with higher precision at the CCSD(T) level with the same basis sets. Based on the obtained results at the DFT level, the energetic parameters are determined to elucidate thermodynamic stability of B_nCu clusters. NICS and ELF indices, the density of states, and AdNDP bond analysis are then examined to rationalize the aromaticity of the anion B₈Cu^-. All calculations were performed using the Gaussian 09 and Multiwfn programs.