Modelling Photocatalytic CO₂ Reduction under Realistic Light and Surface Conditions: A Pathway to Selectivity Control

This study presents a theoretical modeling framework for photocatalytic CO₂ reduction in a cylindrical packed-bed photoreactor, integrating optical simulation, surface activation dy-namics, and kinetic analysis. A ray-based Beer–Lambert model was employed to simulate photon attenuation, revealing a strong vertical gradient in light intensity from 74.43 W/m² at the top to 3.29 W/m² at the bottom underscoring the depth-dependence of catalytic activa-tion. A Weibull-form probability density function (PDF(t)) was used to characterize surface accessibility for methane (CH₄), carbon monoxide (CO), and hydrogen (H₂) pathways, with H₂ showing rapid deactivation and CO maintaining long-term surface activity. Coupling PDF(t) with Langmuir–Hinshelwood kinetics enabled time-resolved reaction rate predic-tions, showing early CH₄ dominance followed by a crossover where CO became the major product after ~10.3 hours. Cumulative yield analysis confirmed this trend, with CO overtak-ing CH₄ at 30 hours and reaching 30.1 µmol/g by 50 hours, while H₂ remained consistently low (~5 µmol/g). The model results aligned well with benchmark experimental data. Sensi-tivity analyses demonstrated that increasing lamp number improved light distribution and CH₄ yield, though with diminishing returns due to deactivation effects. Furthermore, Raschig ring catalyst geometries outperformed spherical pellets by sustaining photon access and surface activation over longer durations. These findings highlight the critical role of re-actor optics and catalyst morphology in determining product selectivity and long-term effi-ciency. The integrated model offers a robust foundation for optimizing photocatalytic sys-tems for sustainable CO₂ utilization.