Machine Learning Assists in Decoupling the Effects of Plasma-Catalytic Ammonia Synthesis

X. Zeng,S. Zhang,X. Hu,T. Shao
DOI: https://doi.org/10.1109/icops58192.2024.10627651
2024-01-01
Abstract:The widespread use of clean, renewable electricity from a wide range of sources, such as wind energy, hydropower, and solar power, has greatly reduced the use of fossil fuels and carbon dioxide emissions [1] . However, this renewable power is not stable enough and needs to be absorbed in place to reduce the impact on the grid [2] . Plasma catalytic ammonia synthesis is an emergently alternative solution, which can produce ammonia, an important industrial raw material, in situ and on a small scale under mild conditions [3] . In this work, a dielectric barrier discharge (DBD) reactor is used to achieve plasma catalytic ammonia synthesis. To decouple the effects of catalyst and plasma on ammonia synthesis, the conversion and energy efficiency of ammonia can be controlled by optimizing the electrical parameters (voltage, frequency, pulse width, etc.) and the loaded metal of the catalyst. Optical diagnosis and fluid simulation reveal the space-time evolution of active species and electric fields to explore the mechanism. The reaction processes of the gas phase and catalyst surface are obtained by introducing the diagnostic parameters into the kinetic model [4] . The experiment shows that there is obvious synergistic effect between plasma and catalyst to promote ammonia synthesis. The machine learning method is also used to analyze the main path of different plasma parameters affecting the ammonia synthesis reaction, which provides an important reference for us to analyze the relationship between the route of nitrogen in NH 3 molecules and the plasma parameters.
What problem does this paper attempt to address?