Adaptive Neuro-Fuzzy inference system analysis on sorption studies of strontium and cesium cations onto a novel impregnated nano-zeolite

Abstract

In this research, a novel impregnated nano-zeolite (NAASMS-Z) was synthesized and characterized using different characterization techniques. Excellent properties, such as high specific surface area (~502.77 m2/g), low pore size (~8.92 Å) and the existence of numerous functional groups caused the efficient elimination of Sr2+ and Cs+ cations from aquatic systems. The sorption performance of the nano-particles was enhanced by impregnation up to 60% in the aquatic media. The kinetic study indicated that the elimination process of both the concerned cations is controlled by external film mass transfer through the boundary within the first 30 min then controlled by intra-particle diffusion. The sorption equilibrium data suggested that the sorption process occurs on the heterogeneous sorbent surface. Parameters affecting the elimination of Sr2+ and Cs+ from a single metal sorption system, such as pH, initial contaminant concentration (Ci) and contact time (t), were investigated and optimized. A predictive model based on an Adaptive Neuro-Fuzzy Inference system (ANFIS) analysis was applied to evaluate the experimental parameters affecting the elimination of Sr2+ and Cs+ cations from aquatic system. A Mamdani-type FIS was employed to justify a collection of 16 rules (If-Then format) by means of centroid membership functions. The suggested fuzzy model revealed high predictive concert with high correlation coefficient (R2) and satisfactory deviation from the experimental data, affirming its appropriateness to predict Sr2+ and Cs+ elimination efficacy from the studied system. Rooted in experimental data and statistical analysis, the synthetized material was effective for treating contaminated aquatic solutions containing Sr2+ and Cs+ cations.