Exploring the potential of SnO2 nanoparticles for CO2 capture using RSM and ANN

This study investigates the properties of SnO2 nanomaterials as CO2 adsorbents. Although CO2 capture technologies have significantly advanced, challenges such as high costs and limited scalability persist. In recent years, nanomaterials have emerged as promising candidates for CO2 adsorption due to their high adsorption capacity, lower cost, and wide availability. However, future research should focus on developing low-cost and efficient nanomaterials to enable large-scale industrial CO2 capture. This study examined the CO2 adsorption capacity using SnO2 nanoadsorbents by employing response surface methodology (RSM) and artificial neural networks (ANNs), specifically radial basis function (RBF) and multilayer perceptron (MLP) networks, for process modeling and optimization. Through the analysis of experimental data, temperature, pressure, and adsorption time were identified as crucial influencing factors. The R2 value of 0.9933 for RSM indicated a great match, whereas the R2 value of one for ANNs indicated superior predictive accuracy. With a minimum mean squared error (MSE) of 0.00012 for the dataset, the MLP was trained using a three-layer activation function. With 288 neurons and a spread of 2, the RBF network achieved an R2 value of 0.9985 and a minimal MSE of 0.00075. The smooth MLP plots effectively captured complex discontinuities, showcasing the superior predictive abilities of ANNs for optimizing the CO2 adsorption process using SnO2 nanoadsorbents, while the RSM surfaces exhibited rigid, polynomial-based patterns.