Isolating the slow component in OSL signals: a numerical approach with neural networks for continuous-field generalization

Pilco E.J.; Cano N.F.; Mosqueira-Yauri J.; Ayala-Arenas J.S.; Benavente J.F. · Radiation Physics and Chemistry (2026) · DOI: 10.1016/j.radphyschem.2026.114036

Resumen

This paper presents a numerical study of a material's response to external stimuli, focusing on an experimentally observed feature of optically stimulated luminescence (OSL) characterized by fast, medium, and slow components. A mathematical model based on a system of ordinary differential equations is developed to describe the time evolution of charge carrier concentrations across the material's energy levels. The numerical framework reproduces experimental procedures used to isolate the slow component and enables detailed analysis of the material's response. The results are formulated in terms of continuous variables, allowing representation as continuous fields and extension to surface-based analysis. Neural networks are employed to reconstruct three-dimensional solution surfaces, accurately capturing the parametric dependence of the response. The model also elucidates limitations of thermal cleaning treatments. Due to intrinsic coupling between component filling and emptying processes, thermal cleaning leads to dominance of the slow component in the luminescence kinetics. This behaviour indicates that thermal cleaning reflects the interplay between thermal and optical processes contributing to the OSL signal.

Instituciones: Universidad Nacional de San Agustín de Arequipa