System-generated electromagnetic pulse (SGEMP) effects induced by X-ray irradiation pose a significant threat to electronic systems in aerospace and nuclear environments. Accurate quantification of electron emission parameters, which are critical current sources for SGEMP simulations, remains challenging because of the complex coupled photon-electron transport processes involved.

This study aims to systematically investigate the characteristics of backward- and forward-emitted electrons from typical materials (e.g., aluminum) under X-ray irradiation and develop efficient analytical models for predicting electron yields without relying on computationally intensive Monte Carlo (MC) simulations for each new scenario.

Photon-electron coupled transport simulations were performed using a Monte Carlo module combining the condensed history and single-event methods. The energy and angular distributions of emitted electrons were analyzed for X rays (0.1–100 keV) normally incident on aluminum plates of varying thicknesses. Analytical models for backward and forward electron yields were derived based on photon mean free path, electron maximum range, and attenuation laws, with a cumulative correction factor introduced to improve forward yield accuracy.

Backward electron energy spectra exhibited a double-peak structure (Compton and photoelectron peaks), with angular distributions following a cosine law. A saturation thickness of about 3 photon mean free paths was identified for backward yield, beyond which yields remained constant. For forward emission, yields peaked at the electron maximum range thickness and decreased with further increasing plate thickness. The proposed analytical formulas for both backward and forward yields achieved relative errors within 10% compared to direct MC simulations across the studied energy and thickness ranges.

The derived analytical models provide efficient and accurate predictions of electron emission coefficients for SGEMP source terms, reducing the need for repeated MC simulations. The methodology is generalizable to other materials and supports rapid assessment of X-ray-induced electron emission in complex systems. Future work will explore machine learning techniques to further enhance computational efficiency for broader applications.