The urgent need for efficient, cost-effective, and regenerable materials for CO2 capture is key to addressing climate change. 

A novel hydrogel composed of polyacrylamide, alpha-olefin sulfonate, and chromium (III) acetate was synthesized and 

investigated for its CO2 adsorption performance for multiple cycles. The inclusion of AOS enhanced gel’s surface activity

and porosity, while Cr(III) acetate served as cross-linker, providing structural stability and active CO2 binding sites. 

Regeneration was achieved by mild heating at 50–60 °C, and the material was reused for 5 cycles, retaining a CO2 adsorption

capacity of 0.60 mmol g⁻¹. Hydrogel maintained excellent mechanical integrity and adsorption capacity, highlighting

its suitability for sustainable CO2 capture applications. To further enhance performance, an artificial neural network 

(ANN) based optimization framework was employed to identify optimal gel compositions. The predictive capabilities of 

Levenberg-Marquardt (LM), particle swarm optimization (PSO), and genetic algorithm (GA) assisted ANN models were 

compared. Among these, the ANN-PSO model exhibited superior predictive accuracy, achieving a correlation coefficient 

of 89.73%, outperforming ANN-GA (84.16%) and ANN-LM (75.77%). The strong agreement between experimental and 

predicted results highlights the effectiveness of ANN-PSO as a powerful tool for optimizing hydrogel-based CO2 capture 

systems.