CO2 dissolution enhancement in seawater by CMC-stabilized nickel nanoparticles

Introduction Improving carbon dioxide (CO 2 ) dissolution in seawater is an important challenge for developing practical marine carbon sequestration technologies, as seawater’shigh salinity limits gas (CO 2 ) dissolution and mass transfer efficiency. In this study, nickel nanoparticles (NiNPs) stabilized with carboxymethyl cellulose (CMC) were investigated to enhance CO2 dissolution in seawater under controlled microfluidic conditions. Methods A flow-focusing microfluidic device was used to generate monodisperse CO 2 microbubbles, and the dissolution behavior was quantified by tracking the reduction in bubble size along the microchannel for different combinations of NiNPs (6-150 mg·L -1 ) and CMC (100-300 mg·L -1 ) concentrations. Results and Discussion The results show that CO 2 dissolution does not increase monotonically with nanoparticle loading but instead follows a concentrationdependent optimum governed by colloidal stability. At low CMC concentration (100 mg·L -1 ), the highest dissolution (∼57%) was observed at 10 mg·L -1 NiNPs, whereas higher nanoparticle loadings reduced performance consistent with reduced colloidal stability and an increased tendency for aggregation in a saline environment. Increasing the CMC concentration improved electrosteric stabilization and shifted the optimal nanoparticle concentration toward higher values, with the maximum dissolution (∼67.2%) observed at 90 mg·L -1 and 300 mg·L -1 CMC. Zeta potential measurements indicated that improved colloidal stability was associated with enhanced dissolution performance, consistent with stability predictions based on DLVO theory. Because the dissolution process depends on multiple interacting physicochemical parameters, an artificial intelligence-based Gaussian Process Regression (GPR) model was developed to interpret nonlinear trends and identify optimal operating conditions. The model demonstrated robust predictive accuracy, achieving a mean relative error of 2.65% against six independent experimental validation points not used in model training, successfully capturing the coupled effects of nanoparticle loading and polymer concentration on CO 2 dissolution. This combined experimental and machine learning approach provides mechanistic insight into carbon capture in saline systems and offers a systematic pathway to optimize marine CO 2 sequestration processes while reducing the need for extensive experimental screening.