Fundamental invariant-neural network potential energy surface and dissociative chemisorption dynamics of N2 on rigid Ni(1 1 1)

Abstract

We constructed a new full-dimensional potential energy surface (PES) for the dissociative chemisorption of N2 on rigid Ni(111). The PES was fitted using the fundamental invariant-neural network method based on DFT energy points. The resulting PES is accurately fitted with very small fitting error of about 4.5 meV, which reproduces well the minimum energy path and energies of stationary points. Based on the new PES, the effects of incidence energy, incidence angle as well as the vibrational excitations of impinging molecules on dissociation probability were investigated by quasi-classical trajectory calculations. The vibrational excitation enhances the reactivity more effectively than the same amount of translational energy, which is consistent with the very “product-like” transition state for this system. The dissociation occurs close to the bridge site, presumably due to the extremely tight and high-energy barriers of PES at top and hollow sites.