Analysis of the Gas Phase Kinetics Active during GaN Deposition from NH3 and Ga(CH3)3

The results of a systematic investigation aimed at determining the dominant gas phase chemistry active during GaN MOVPE are reported and discussed in this work. This study was performed developing a thermodynamic database including the most stable GaN gas phase species and a gas phase mechanism that could efficiently describe their interconversion kinetics. The thermodynamic data and the kinetic mechanism were calculated combining density functional theory and ab initio simulations. Structures and vibrational frequencies of reactants and transition states were determined at the M062X/6-311+G(d,p) level, while energies were computed at the ROCBS-QB3 level. Rate constants were calculated using transition state theory using the rigid rotor - harmonic oscillator approximation and considering the possible degeneration of internal motions in torsional rotations. The thermodynamic analysis indicated that the Ga gas phase species formed in the highest concentration at the standard GaN deposition temperature (1300 K) is GaNH2, followed by GaH and Ga. The diatomic GaN gas phase species, often considered to be the main precursor to the film growth, is predicted to be unstable with respect to GaNH2. Among the gas phase species containing two Ga atoms, the most stable are GaNHGaH(NH2)3, GaNHGaH2(NH2)2, and GaNHGa(NH2)4, thus indicating that the substitution of the methyl groups of the precursor with H or amino groups is thermodynamically favored. Several kinetic routes leading to the formation of these species were examined. It was found that the condensation of Ga(R1)x(R2)3-x species, with R1 and R2 being either CH3, NH2, or H, is a fast process, characterized by the formation of a precursor state whose decomposition to products requires overcoming submerged energy barriers. It is suggested that these species play a key role in the formation of the first GaN nuclei, whose successive growth leads to the formation of GaN powders. A kinetic analysis performed using a fluid dynamic model allowed us to identify the main reactive routes of this complex system.