Structural, electronic and vibrational properties of small Ga(x)N(y) (x+y = 2-5) nanoclusters: a B3LYP-DFT study

An ab initio study of the stability, structural and electronic properties has been made for 49 gallium nitride nanoclusters, Ga(x)N(y) (x+y = 2-5). Among the various configurations corresponding to a fixed x+y = n value, the configuration possessing the maximum value of binding energy (BE) is named as the most stable structure. The vibrational and optical properties have been investigated only for the most stable structures. A B3LYP-DFT/6-311G(3df) method has been employed to optimize the geometries of the nanoclusters fully. The binding energies (BEs), highest-occupied and lowest-unoccupied molecular orbital (HOMO-LUMO) gaps and the bond lengths have been obtained for all the clusters. We have considered the zero-point energy (ZPE) corrections ignored by the earlier workers. The adiabatic and vertical ionization potentials (IPs) and electron affinities (EAs), charge on atoms, dipole moments, vibrational frequencies, infrared intensities (IR Int.), relative infrared intensities (Rel. IR Int.) and Raman scattering activities have been investigated for the most stable structures. The configurations containing the N atoms in majority are seen to be the most stable structures. The strong N-N bond has an important role in stabilizing the clusters. For clusters containing one Ga atom and all the others as N atoms, the BE increases monotonically with the number of the N atoms. The HOMO-LUMO gap and IP fluctuate with the cluster size n, having larger values for the clusters containing odd number of N atoms. On the other hand, the EA decreases with the cluster size up to n = 3, and shows slow fluctuations thereafter for the larger clusters. In general, the adiabatic IP (EA) is smaller (greater) than the vertical IP (EA) because of the lower energies of the most stable ground state of the cationic (anionic) clusters. The optical absorption spectrum or electron energy loss spectrum (EELS) is unique for every cluster, and may be used to characterize a specific cluster. All the predicted physical quantities are in good agreement with the experimental data wherever available. The growth of these most stable structures should be possible in experiments.