PHITS–ANN investigation of gamma and neutron shielding performance of recycled glass–cement composites reinforced with metal organic frameworks: Bi₂O₃, and B₄C

This research investigates the radiation shielding performance of recycled glass–cement composites doped with Bi₂O₃ and B₄C and reinforced with metal–organic frameworks (MOFs) using PHITS Monte Carlo simulations combined with an artificial neural network (ANN) predictive model. The incorporation of MOFs is explored as a novel strategy to enhance both gamma-ray attenuation and neutron moderation, owing to their high surface area, tunable porosity, and ability to host high-Z metal centers. In addition to their shielding contribution, MOFs may improve the structural homogeneity of the composite matrix, offering multifunctional advantages. The interaction of photons and neutrons with the developed composites is evaluated in terms of the linear attenuation coefficient, mass attenuation coefficient, half-value layer, and effective removal cross-section over a wide energy range. The PHITS simulation results indicate that the synergistic combination of Bi₂O₃, B₄C, and MOFs significantly improves radiation shielding efficiency compared to conventional glass–cement matrices. Furthermore, the ANN model is trained on the simulated dataset to predict shielding parameters with high accuracy, reducing computational cost and enabling rapid optimization of material composition. The integration of simulation and machine learning provides an efficient framework for accelerated material design and performance prediction. The results demonstrate that MOF-reinforced recycled glass–cement composites exhibit promising multifunctional properties, making them strong candidates for advanced, sustainable shielding materials in nuclear technology, medical radiation protection, and other radiation-intensive applications.