CuO nanoparticle doping in barium borate glasses: effects on density prediction via machine learning, optical, mechanical, thermal, and radiation protection properties

Abstract This study synthesized CuO nanoparticles (average size 2.75 nm) via combustion and doped them into a series of melt-quenched borate glasses with the composition (70- x )B₂O₃: x CuO:10CaO:10 K₂O:5Na₂O:5BaO ( x = 0, 0.25, 0.5, 0.75, 1 mol%). CuO addition increased the thermal stability, density (from 2.513 to 2.535), and refractive index (from 2.26 to 2.35) of the glasses. The experimental density values were validated using machine learning models, with the XGB Reg. model ( R 2 = 0.986) providing the most accurate predictions. FTIR spectroscopy identified vibrational modes such as B–O–B bending (718 cm⁻ 1 ), BO₄ unit vibrations (1054 cm⁻ 1 ), and B–O stretching (1360 cm⁻ 1 ), as well as non-bridging oxygen sites influenced by CuO content. Optical analysis revealed a decrease in both direct and indirect bandgap energies, an increase in refractive index, and improved functionality as absorption bandpass filters and UV shields. Mechanical evaluation using the Makishima–Mackenzie model showed a reduction in elastic modulus values with increasing CuO content, linked to lower packing density. Thermal analysis confirmed consistent stability and glass-forming ability. Radiation shielding tests using γ-rays (384, 1173, 1333 keV) showed improved efficiency with increasing CuO, highlighting the potential for developing advanced radiation shielding materials.