Unveiling state-specific total-exchange dynamics in high-temperature N2–N2/N2–O2 collisions via QCT and neural-network modeling

Abstract

State-specific molecular–molecular exchange (total-exchange) reactions are pivotal microscopic processes governing vibrational energy redistribution and dissociation kinetics in high-temperature non-equilibrium gases. Quasi-classical trajectory (QCT) calculations are performed to elucidate the competition among dissociation and exchange channels in N2(v1) + N2(v2) and N2(v) + O2(w) collisions. The results reveal that under low vibrational excitation (v ≤ 20), total-exchange emerges as the dominant reaction pathway in N2–N2 collisions, surpassing all dissociation channels. In the N2–O2 system, although total-exchange is generally a secondary pathway, its contribution can surpass that of the typically dominant O2-dissoc when highly excited N2 collides with ground-state O2. Notably, in both systems, the total-exchange cross section increases monotonically with total collision energy below the double-dissociation threshold, and its high-temperature rate coefficients are comparable to those of non-reactive vibration–vibration and vibration–translation (VV/VT) energy transfer, underscoring its critical role in driving multi-quantum vibrational transitions. To enable efficient large-scale prediction, we develop two neural-network models (N4-NN, N2O2-NN) trained on the QCT data that achieve excellent accuracy (R2 > 0.99) with ∼91% lower computational cost. The predicted cross sections are fitted to a compact analytical form, yielding a parameterized database ready for engineering applications. This study elucidates the state-specific mechanism of total-exchange-controlled energy transfer and its energy-dependent evolution, providing a comprehensive data foundation for modeling high-temperature non-equilibrium flows.