AI-driven parametric optimization of gas-liquid absorption for the intensification of CO2 capture under a Gas-phase pulsation condition

Abstract

Traditional CO₂ capture using amine-based solvents is effective but not an energy-intensive and requires frequent replenishment. This study explores enhancing CO₂ absorption in packed bed columns by switching to sodium hydroxide and incorporating gas phase pulsation to improve mass transfer efficiency. Optimising a CO₂NaOH absorption process through its intensified volumetric mass transfer coefficient under a gas phase pulsation using artificial intelligence model is the main objective of this study. The mutual effects of pulsation amplitude, frequency, bed height, and solvent content on volumetric mass transfer coefficient was observed by Central Composite Design model of Response Surface Methodology where under an ideal frequency of 7.5 Hz, an amplitude of 18 mm, a bed height of 12 cm, and a solvent concentration of 2 N, the model attained a maximum volumetric mass transfer coefficient of 53.166 ± 0.55 s^-1. This result was further validated through the Genetic Algorithm and Particle Swarm Optimisation models of Artificial Neural Networks. It revealed maximum coefficients of 54.52 ± 40 s^-1 and 56.12 ± 60 s^-1, respectively, with marginally differing ideal parameters. This study shows that artificial intelligence can substantially optimize CO₂ capture processes by maximizing the volumetric mass transfer coefficient, leading to more efficient and cost-effective greenhouse gas reduction methods.