Toward a complete and comprehensive cross section database for electron scattering from NO using machine learning

Area of Science:

• Atomic and Molecular Physics
• Plasma Physics
• Computational Physics

Background:

• Accurate electron scattering cross sections are crucial for understanding electron transport in gases like nitric oxide (NO).
• Previous cross section sets for NO showed discrepancies with experimental swarm data, necessitating refinement.

Purpose of the Study:

• To develop a comprehensive and accurate set of electron scattering cross sections for nitric oxide (NO).
• To improve the agreement between theoretical calculations and experimental measurements of electron transport in NO.

Main Methods:

• Review of experimental and theoretical electron scattering cross sections for NO.
• Multi-term Boltzmann equation analysis of electron swarm transport coefficients in pure NO and NO/Ar mixtures.
• Artificial neural network training to refine cross sections by solving the inverse problem against swarm data.

Main Results:

• A refined set of electron-NO cross sections, including quasielastic momentum transfer, dissociative attachment, and neutral dissociation.
• Demonstrated improved agreement of the refined cross sections with experimental electron swarm transport data.
• Calculation of electron transport coefficients in NO across a wide range of reduced electric fields (0.003–10,000 Td).

Conclusions:

• The refined electron-NO cross section set provides a more accurate representation for electron transport studies.
• Artificial neural network-based refinement is an effective method for improving cross section databases.
• The study provides a reliable dataset for electron-NO interactions relevant to various applications.